Introduce a new StreamingEditFileTool (#29733)

This pull request introduces a new tool for streaming edits. The
short-term goal is for this tool to replace the existing `EditFileTool`,
but we want to get this out the door as soon as possible so that we can
start testing it.

`StreamingEditFileTool` is mutually exclusive with `EditFileTool`. It
will be enabled by default for anyone who has the `agent-stream-edits`
feature flag, as well as people that set `assistant.stream_edits` to
`true` in their settings.

### Implementation

Streaming is achieved by requesting a completion while the `edit_file`
tool gets called. We invoke the model by taking the existing
conversation with the agent and appending a prompt specifically tailored
for editing. In that prompt, we ask the model to produce a stream of
`<old_text>`/`<new_text>` tags. As the model streams text in, we
incrementally parse it and start editing as soon as we can.

### Evals

Note that, as part of this pull request, I also defined some new evals
that I used to drive the behavior of the recursive LLM call. To run
them, use this command:

```bash
cargo test --package=assistant_tools --features eval -- eval_extract_handle_command_output
```

Or comment out the `#[cfg_attr(not(feature = "eval"), ignore)]` macro.

I recommend running them one at a time, because right now we don't
really have a way of orchestrating of all these evals. I think we should
invest into that effort once the new agent panel goes live.

Release Notes:

- N/A

---------

Co-authored-by: Nathan Sobo <nathan@zed.dev>
Co-authored-by: Bennet Bo Fenner <bennetbo@gmx.de>
Co-authored-by: Oleksiy Syvokon <oleksiy.syvokon@gmail.com>

crates/eval/Cargo.toml

@@ -48,6 +48,7 @@ markdown.workspace = true
 node_runtime.workspace = true
 pathdiff.workspace = true
 paths.workspace = true
+pretty_assertions.workspace = true
 project.workspace = true
 prompt_store.workspace = true
 regex.workspace = true

crates/eval/runner_settings.json

@@ -1,6 +1,7 @@
 {
   "assistant": {
     "always_allow_tool_actions": true,
+    "stream_edits": true,
     "version": "2"
   }
 }

crates/eval/src/eval.rs

@@ -420,12 +420,12 @@ pub fn init(cx: &mut App) -> Arc<AgentAppState> {
     language_model::init(client.clone(), cx);
     language_models::init(user_store.clone(), client.clone(), fs.clone(), cx);
     languages::init(languages.clone(), node_runtime.clone(), cx);
-    assistant_tools::init(client.http_client(), cx);
     context_server::init(cx);
     prompt_store::init(cx);
     let stdout_is_a_pty = false;
     let prompt_builder = PromptBuilder::load(fs.clone(), stdout_is_a_pty, cx);
     agent::init(fs.clone(), client.clone(), prompt_builder.clone(), cx);
+    assistant_tools::init(client.http_client(), cx);
 
     SettingsStore::update_global(cx, |store, cx| {
         store.set_user_settings(include_str!("../runner_settings.json"), cx)

crates/eval/src/example.rs

@@ -160,7 +160,11 @@ impl ExampleContext {
             if left == right {
                 Ok(())
             } else {
-                println!("{}{:#?} != {:#?}", self.log_prefix, left, right);
+                println!(
+                    "{}{}",
+                    self.log_prefix,
+                    pretty_assertions::Comparison::new(&left, &right)
+                );
                 Err(anyhow::Error::from(FailedAssertion(message.clone())))
             },
             message,
@@ -334,8 +338,8 @@ impl ExampleContext {
     }
 
     pub fn edits(&self) -> HashMap<Arc<Path>, FileEdits> {
-        self.app
-            .read_entity(&self.agent_thread, |thread, cx| {
+        self.agent_thread
+            .read_with(&self.app, |thread, cx| {
                 let action_log = thread.action_log().read(cx);
                 HashMap::from_iter(action_log.changed_buffers(cx).into_iter().map(
                     |(buffer, diff)| {
@@ -503,16 +507,16 @@ impl ToolUse {
     }
 }
 
-#[derive(Debug)]
+#[derive(Debug, Eq, PartialEq)]
 pub struct FileEdits {
-    hunks: Vec<FileEditHunk>,
+    pub hunks: Vec<FileEditHunk>,
 }
 
-#[derive(Debug)]
-struct FileEditHunk {
-    base_text: String,
-    text: String,
-    status: DiffHunkStatus,
+#[derive(Debug, Eq, PartialEq)]
+pub struct FileEditHunk {
+    pub base_text: String,
+    pub text: String,
+    pub status: DiffHunkStatus,
 }
 
 impl FileEdits {