fix(harness): use real newlines in sandbox edit python -c

agentscope-harness/src/main/java/io/agentscope/harness/agent/filesystem/sandbox/BaseSandboxFilesystem.java

@@ -224,32 +224,43 @@ public EditResult edit(
                 Base64.getEncoder()
                         .encodeToString(payload.getBytes(java.nio.charset.StandardCharsets.UTF_8));
 
+        // Use real newlines inside the python3 -c program. Java "\\n" would emit a
+        // literal backslash-n, which POSIX shells pass through and Python rejects
+        // as a SyntaxError (see #2571).
         String cmd =
-                "python3 -c \"import sys, os, base64, json\\n"
+                "python3 -c \"import sys, os, base64, json\n"
                     + "payload ="
-                    + " json.loads(base64.b64decode(sys.stdin.read().strip()).decode('utf-8'))\\n"
-                    + "path, old, new = payload['path'], payload['old'], payload['new']\\n"
-                    + "replace_all = payload.get('replace_all', False)\\n"
-                    + "if not os.path.isfile(path):\\n"
-                    + "    print(json.dumps({'error': 'file_not_found'}))\\n"
-                    + "    sys.exit(0)\\n"
-                    + "with open(path, 'rb') as f: text = f.read().decode('utf-8')\\n"
-                    + "count = text.count(old)\\n"
-                    + "if count == 0:\\n"
-                    + "    print(json.dumps({'error': 'string_not_found'}))\\n"
-                    + "    sys.exit(0)\\n"
-                    + "if count > 1 and not replace_all:\\n"
-                    + "    print(json.dumps({'error': 'multiple_occurrences', 'count': count}))\\n"
-                    + "    sys.exit(0)\\n"
+                    + " json.loads(base64.b64decode(sys.stdin.read().strip()).decode('utf-8'))\n"
+                    + "path, old, new = payload['path'], payload['old'], payload['new']\n"
+                    + "replace_all = payload.get('replace_all', False)\n"
+                    + "if not os.path.isfile(path):\n"
+                    + "    print(json.dumps({'error': 'file_not_found'}))\n"
+                    + "    sys.exit(0)\n"
+                    + "with open(path, 'rb') as f: text = f.read().decode('utf-8')\n"
+                    + "count = text.count(old)\n"
+                    + "if count == 0:\n"
+                    + "    print(json.dumps({'error': 'string_not_found'}))\n"
+                    + "    sys.exit(0)\n"
+                    + "if count > 1 and not replace_all:\n"
+                    + "    print(json.dumps({'error': 'multiple_occurrences', 'count': count}))\n"
+                    + "    sys.exit(0)\n"
                     + "result = text.replace(old, new) if replace_all else text.replace(old, new,"
-                    + " 1)\\n"
-                    + "with open(path, 'wb') as f: f.write(result.encode('utf-8'))\\n"
-                    + "print(json.dumps({'count': count}))\\n"
+                    + " 1)\n"
+                    + "with open(path, 'wb') as f: f.write(result.encode('utf-8'))\n"
+                    + "print(json.dumps({'count': count}))\n"
                     + "\" 2>&1 <<'__EDIT_EOF__'\n"
                         + payloadB64
                         + "\n__EDIT_EOF__\n";
 
         ExecuteResponse result = execute(runtimeContext, cmd, null);
+        if (result.exitCode() != null && result.exitCode() != 0) {
+            String err = result.output() != null ? result.output().strip() : "";
+            return EditResult.fail(
+                    "Error editing file '"
+                            + filePath
+                            + "': "
+                            + err.substring(0, Math.min(200, err.length())));
+        }
         String output = result.output() != null ? result.output().strip() : "";
 
         if (output.contains("\"error\"")) {

agentscope-harness/src/test/java/io/agentscope/harness/agent/filesystem/sandbox/BaseSandboxFilesystemTest.java

@@ -130,6 +130,72 @@ void ls_reportsDirModifiedAt() {
             assertTrue(dir.isDirectory());
             assertFalse(dir.modifiedAt().isEmpty(), "dir modifiedAt should be populated");
         }
+
+        @Test
+        void edit_pythonCommand_usesRealNewlinesNotLiteralBackslashN() {
+            FakeSandboxFilesystem filesystem = new FakeSandboxFilesystem();
+            filesystem.nextEditResponse = new ExecuteResponse("{\"count\": 1}", 0, false);
+
+            var result = filesystem.edit(RT, "/tmp/example.txt", "old", "new", false);
+
+            assertTrue(result.isSuccess());
+            // Real LF between "json" and "payload" — not the two-char sequence '\' + 'n'.
+            assertTrue(
+                    filesystem.lastCommand.contains("json\npayload"),
+                    "python -c program must contain real newlines");
+            assertFalse(
+                    filesystem.lastCommand.contains("json" + "\\" + "npayload"),
+                    "must not emit literal backslash-n (POSIX shell / Python SyntaxError)");
+        }
+
+        @Test
+        void edit_nonzeroExitCode_returnsFailureWithOutput() {
+            FakeSandboxFilesystem filesystem = new FakeSandboxFilesystem();
+            filesystem.nextEditResponse =
+                    new ExecuteResponse("SyntaxError: unexpected character", 1, false);
+
+            var result = filesystem.edit(RT, "/tmp/example.txt", "old", "new", false);
+
+            assertFalse(result.isSuccess());
+            assertTrue(result.error().contains("SyntaxError"));
+        }
+
+        @Test
+        void edit_nonzeroExitCode_nullOutput_returnsFailure() {
+            FakeSandboxFilesystem filesystem = new FakeSandboxFilesystem();
+            // Covers: exitCode != 0 && output == null -> err = ""
+            filesystem.nextEditResponse = new ExecuteResponse(null, 1, false);
+
+            var result = filesystem.edit(RT, "/tmp/example.txt", "old", "new", false);
+
+            assertFalse(result.isSuccess());
+            assertTrue(result.error().contains("Error editing file"));
+        }
+
+        @Test
+        void edit_nullExitCode_withCount_stillSucceeds() {
+            FakeSandboxFilesystem filesystem = new FakeSandboxFilesystem();
+            // Covers: exitCode == null -> skip nonzero-fail branch
+            filesystem.nextEditResponse = new ExecuteResponse("{\"count\": 2}", null, false);
+
+            var result = filesystem.edit(RT, "/tmp/example.txt", "old", "new", false);
+
+            assertTrue(result.isSuccess());
+            assertEquals(2, result.occurrences());
+        }
+
+        @Test
+        void edit_nonzeroExitCode_longOutput_isTruncatedTo200() {
+            FakeSandboxFilesystem filesystem = new FakeSandboxFilesystem();
+            String longErr = "E".repeat(250);
+            filesystem.nextEditResponse = new ExecuteResponse(longErr, 2, false);
+
+            var result = filesystem.edit(RT, "/tmp/example.txt", "old", "new", false);
+
+            assertFalse(result.isSuccess());
+            assertTrue(result.error().endsWith("E".repeat(200)));
+            assertFalse(result.error().contains("E".repeat(201)));
+        }
     }
 
     // ================================================================
@@ -215,6 +281,7 @@ void glob_emptyResultWhenNoMatch() {
     private static final class FakeSandboxFilesystem extends BaseSandboxFilesystem {
 
         String lastCommand;
+        ExecuteResponse nextEditResponse;
 
         @Override
         public String id() {
@@ -225,6 +292,9 @@ public String id() {
         public ExecuteResponse execute(
                 RuntimeContext runtimeContext, String command, Integer timeoutSeconds) {
             lastCommand = command;
+            if (command.startsWith("python3 -c") && nextEditResponse != null) {
+                return nextEditResponse;
+            }
             if (command.startsWith("for f in ") && command.contains("stat -c")) {
                 return new ExecuteResponse(
                         "DIR:/workspace/docs\t1719300000\n"