feat: 文档/图片工具进任务镜像 + 结果截断从截断 JSON 改为提取 text 上限 20000 字

cloud/tests/test_queue_sandbox.py

@@ -20,7 +20,8 @@ sys.path.insert(0, os.path.dirname(os.path.dirname(
 
 from queue import TaskQueue, budget_exceeded, order_tasks, ready_tasks
 from sandbox import (CloudExecutor, DockerExecutor, Executor,
-                     LocalExecutor, _parse_goal_output)
+                     LocalExecutor, _parse_goal_output,
+                     _parse_host_result, _result_text)
 
 
 def _task(task_id, depends_on=None, priority=0, status="queued"):
@@ -177,12 +178,38 @@ class TestExecutors(unittest.TestCase):
         self.assertEqual(result["stdout"], "boom")
 
     def test_local_executor_truncates_stdout(self):
+        # 20000 字上限：5000 字不再截断，25000 字截断并带标记。
         def fake_runner(cmd):
             return SimpleNamespace(returncode=0, stdout="x" * 5000)
 
         result = LocalExecutor("/fake/host.py",
                                runner=fake_runner).execute({"id": "t3"})
-        self.assertEqual(len(result["stdout"]), 2000)
+        self.assertEqual(len(result["stdout"]), 5000)
+
+        def long_runner(cmd):
+            return SimpleNamespace(returncode=0, stdout="y" * 25000)
+
+        result = LocalExecutor("/fake/host.py",
+                               runner=long_runner).execute({"id": "t4"})
+        self.assertIn("结果已截断", result["stdout"])
+        self.assertTrue(result["stdout"].startswith("y" * 20000))
+
+    def test_result_text_extracts_long_json_text(self):
+        # 2026-10-30 生产审计任务的复发：host.py JSON 的 text
+        # 超过旧 2000 字截断时，存库的不能是截断的 JSON，
+        # 必须是完整提取出的 text（本例 5000 字，全部保留）。
+        text = "z" * 5000
+        payload = json.dumps({"status": "completed", "text": text})
+        result = _parse_host_result(
+            "t5", SimpleNamespace(returncode=0, stdout=payload,
+                                  stderr=""))
+        self.assertEqual(result["stdout"], text)
+        self.assertEqual(result["status"], "completed")
+
+    def test_result_text_marks_over_limit(self):
+        text = "w" * 21000
+        self.assertIn("结果已截断", _result_text(
+            {"text": text}, "unused"))
 
     def test_docker_executor_no_docker_raises_not_local(self):
         with mock.patch("sandbox.shutil.which", return_value=None):