81 lines
3.0 KiB
Python
81 lines
3.0 KiB
Python
|
||
import pytest
|
||
from unittest.mock import MagicMock, patch, AsyncMock
|
||
from src.classes.mutual_action.conversation import Conversation
|
||
from src.classes.action_runtime import ActionStatus
|
||
from src.classes.event import Event
|
||
|
||
class TestActionSocial:
|
||
|
||
@pytest.fixture
|
||
def target_avatar(self, dummy_avatar):
|
||
target = MagicMock()
|
||
target.name = "FriendDummy"
|
||
target.id = "friend_id"
|
||
target.get_info.return_value = "Target Info"
|
||
target.get_planned_actions_str.return_value = "None"
|
||
target.thinking = ""
|
||
# 模拟 add_event
|
||
target.events = []
|
||
target.add_event = lambda e, to_sidebar=False: target.events.append(e)
|
||
# 模拟修炼进度（用于关系判断）
|
||
target.cultivation_progress.level = 10
|
||
target.gender = dummy_avatar.gender # 同性
|
||
target.get_relation.return_value = None
|
||
|
||
return target
|
||
|
||
@pytest.mark.asyncio
|
||
@patch("src.classes.mutual_action.mutual_action.call_llm_with_template", new_callable=AsyncMock)
|
||
async def test_conversation_flow(self, mock_llm, dummy_avatar, target_avatar):
|
||
"""测试对话流程：Step -> LLM -> Feedback"""
|
||
|
||
# 1. 准备 Mock LLM 返回
|
||
mock_response = {
|
||
"FriendDummy": {
|
||
"thinking": "He is nice.",
|
||
"conversation_content": "Hello there!",
|
||
"feedback": "Accept" # Conversation 其实不强制 feedback，主要是 content
|
||
}
|
||
}
|
||
mock_llm.return_value = mock_response
|
||
|
||
# 注入 World 查找
|
||
dummy_avatar.world.avatar_manager.avatars = {target_avatar.name: target_avatar}
|
||
|
||
# Mock 自己的 level (避免 dummy_avatar 中也是 Mock 导致无法比较)
|
||
dummy_avatar.cultivation_progress.level = 10
|
||
|
||
# 2. 初始化 Action
|
||
action = Conversation(dummy_avatar, dummy_avatar.world)
|
||
action._start_month_stamp = 100
|
||
|
||
# 3. 第一次 Step: 应该触发 LLM 任务并返回 RUNNING
|
||
res1 = action.step(target_avatar=target_avatar)
|
||
assert res1.status == ActionStatus.RUNNING
|
||
assert action._feedback_task is not None
|
||
|
||
# 等待 Task 完成
|
||
await action._feedback_task
|
||
|
||
# 4. 第二次 Step: 消费结果
|
||
res2 = action.step(target_avatar=target_avatar)
|
||
assert res2.status == ActionStatus.COMPLETED
|
||
|
||
# 5. 验证结果
|
||
# 应该有一个包含对话内容的事件
|
||
assert len(res2.events) >= 1
|
||
content_event = res2.events[0]
|
||
assert "Hello there!" in content_event.content
|
||
assert dummy_avatar.id in content_event.related_avatars
|
||
assert target_avatar.id in content_event.related_avatars
|
||
|
||
# 验证 Target 思考被更新
|
||
assert target_avatar.thinking == "He is nice."
|
||
|
||
def test_conversation_no_target(self, dummy_avatar):
|
||
action = Conversation(dummy_avatar, dummy_avatar.world)
|
||
res = action.step(target_avatar=None)
|
||
assert res.status == ActionStatus.FAILED
|
||
|