Prevent invalid utf8 indexing in cell magic detection (#9146)

The example below used to panic because we tried to split at 2 bytes in
the 4-bytes character `转`.
```python
def sample_func(xx):
    """
    转置 (transpose)
    """
    return xx.T
```

Fixes #9145
Fixes https://github.com/astral-sh/ruff-vscode/issues/362

The second commit is a small test refactoring.

crates/ruff_notebook/resources/test/fixtures/jupyter/cell/unicode_magic_gh9145.json

@@ -0,0 +1,19 @@
+{
+  "execution_count": null,
+  "cell_type": "code",
+  "id": "1",
+  "metadata": {},
+  "outputs": [],
+  "source": [
+    "def sample_func(xx):\n",
+    "    \"\"\"\n",
+    "    转置 (transpose)\n",
+    "    \"\"\"\n",
+    "    return xx.T",
+    "# https://github.com/astral-sh/ruff-vscode/issues/362",
+    "DEFAULT_SYSTEM_PROMPT = (",
+    "    \"Ты — Сайга, русскоязычный автоматический ассистент. \"",
+    "    \"Ты разговариваешь с людьми и помогаешь им.\"",
+    ")"
+  ]
+}

crates/ruff_notebook/src/cell.rs

@@ -171,48 +171,43 @@ impl Cell {
 
         // Detect cell magics (which operate on multiple lines).
         lines.any(|line| {
-            line.split_whitespace().next().is_some_and(|first| {
+            let Some(first) = line.split_whitespace().next() else {
-                if first.len() < 2 {
+                return false;
-                    return false;
+            };
-                }
+            if first.len() < 2 {
-                let (token, command) = first.split_at(2);
+                return false;
-                // These cell magics are special in that the lines following them are valid
+            }
-                // Python code and the variables defined in that scope are available to the
+            let Some(command) = first.strip_prefix("%%") else {
-                // rest of the notebook.
+                return false;
-                //
+            };
-                // For example:
+            // These cell magics are special in that the lines following them are valid
-                //
+            // Python code and the variables defined in that scope are available to the
-                // Cell 1:
+            // rest of the notebook.
-                // ```python
+            //
-                // x = 1
+            // For example:
-                // ```
+            //
-                //
+            // Cell 1:
-                // Cell 2:
+            // ```python
-                // ```python
+            // x = 1
-                // %%time
+            // ```
-                // y = x
+            //
-                // ```
+            // Cell 2:
-                //
+            // ```python
-                // Cell 3:
+            // %%time
-                // ```python
+            // y = x
-                // print(y)  # Here, `y` is available.
+            // ```
-                // ```
+            //
-                //
+            // Cell 3:
-                // This is to avoid false positives when these variables are referenced
+            // ```python
-                // elsewhere in the notebook.
+            // print(y)  # Here, `y` is available.
-                token == "%%"
+            // ```
-                    && !matches!(
+            //
-                        command,
+            // This is to avoid false positives when these variables are referenced
-                        "capture"
+            // elsewhere in the notebook.
-                            | "debug"
+            !matches!(
-                            | "prun"
+                command,
-                            | "pypy"
+                "capture" | "debug" | "prun" | "pypy" | "python" | "python3" | "time" | "timeit"
-                            | "python"
+            )
-                            | "python3"
-                            | "time"
-                            | "timeit"
-                    )
-            })
         })
     }
 }

crates/ruff_notebook/src/notebook.rs

@@ -421,17 +421,18 @@ mod tests {
         ));
     }
 
-    #[test_case(Path::new("markdown.json"), false; "markdown")]
+    #[test_case("markdown", false)]
-    #[test_case(Path::new("only_magic.json"), true; "only_magic")]
+    #[test_case("only_magic", true)]
-    #[test_case(Path::new("code_and_magic.json"), true; "code_and_magic")]
+    #[test_case("code_and_magic", true)]
-    #[test_case(Path::new("only_code.json"), true; "only_code")]
+    #[test_case("only_code", true)]
-    #[test_case(Path::new("cell_magic.json"), false; "cell_magic")]
+    #[test_case("cell_magic", false)]
-    #[test_case(Path::new("valid_cell_magic.json"), true; "valid_cell_magic")]
+    #[test_case("valid_cell_magic", true)]
-    #[test_case(Path::new("automagic.json"), false; "automagic")]
+    #[test_case("automagic", false)]
-    #[test_case(Path::new("automagics.json"), false; "automagics")]
+    #[test_case("automagics", false)]
-    #[test_case(Path::new("automagic_before_code.json"), false; "automagic_before_code")]
+    #[test_case("automagic_before_code", false)]
-    #[test_case(Path::new("automagic_after_code.json"), true; "automagic_after_code")]
+    #[test_case("automagic_after_code", true)]
-    fn test_is_valid_code_cell(path: &Path, expected: bool) -> Result<()> {
+    #[test_case("unicode_magic_gh9145", true)]
+    fn test_is_valid_code_cell(cell: &str, expected: bool) -> Result<()> {
         /// Read a Jupyter cell from the `resources/test/fixtures/jupyter/cell` directory.
         fn read_jupyter_cell(path: impl AsRef<Path>) -> Result<Cell> {
             let path = notebook_path("cell").join(path);
@@ -439,7 +440,10 @@ mod tests {
             Ok(serde_json::from_str(&source_code)?)
         }
 
-        assert_eq!(read_jupyter_cell(path)?.is_valid_code_cell(), expected);
+        assert_eq!(
+            read_jupyter_cell(format!("{cell}.json"))?.is_valid_code_cell(),
+            expected
+        );
         Ok(())
     }