edit_file: Add diff-fenced output format (#32737)

This format is enabled for Google models as they seem to prefer it.
A relevant unit eval's pass rate has increased from 0.77 to 0.98.

Diff-fenced format looks like this (markdown fences and a line hint are
optional):

```diff
<<<<<<< SEARCH line=42
...
=======
...
>>>>>>> REPLACE
```

Release Notes:

- Agent: Gemini models now use the diff-fenced format when making edits

crates/assistant_tools/src/edit_agent/streaming_fuzzy_matcher.rs

@@ -44,6 +44,10 @@ impl StreamingFuzzyMatcher {
     /// Returns `Some(range)` if a match has been found with the accumulated
     /// query so far, or `None` if no suitable match exists yet.
     pub fn push(&mut self, chunk: &str, line_hint: Option<u32>) -> Option<Range<usize>> {
+        if line_hint.is_some() {
+            self.line_hint = line_hint;
+        }
+
         // Add the chunk to our incomplete line buffer
         self.incomplete_line.push_str(chunk);
         self.line_hint = line_hint;