fix conflict

tests/data/embedding/2.answer.md

@@ -0,0 +1,2 @@
+检索结果
+法务查询者可根据国际小超人钉钉小程序UI上的滚筒切换业务线  这张图片展示了一个移动应用的界面，界面标题为“法律意见详情”。用户可以根据具体情况切换业务线。界面中有多个字段，包括“国家名称”、“国家情况描述”、“业务线”、“产品法规分析”和“签约主体”。第一张截图显示了详细的法律情报信息，包含区域名称、区域情况描述、业务线和产品法规概述等字段。第二张截图显示了“法律意见详情”界面，其中列出了国家名称、国家情况描述、业务线、产品法规分析和签约主体。第三张截图与第二张相似，但显示了选项的可选择状态。最下方有“取消”和“确定”的按钮。 法务查询者从国家详情中的业务线名列表中选出要查看的业务线。

tests/data/embedding/2.knowledge.md

@@ -0,0 +1,25 @@
+## Textual User Requirements
+
+### 3.2. 首页
+
+首页有两个分区，上面部分是法律意见检索栏。
+
+法务查询者第一次进入国际小超人钉钉小程序展示引导页，以后进入不再展示，点击「我知道了」引导页消失。
+
+#### 首页
+![首页](1.png)
+这是一个名为“法务小超人”的移动应用程序的界面截图。界面顶部显示了应用名称和一个可切换语言的按钮“English”。在界面中间部分，有一个标题“法律意见查询”，以及一个搜索框，提示输入国家名称以查询法律意见。下方显示已收录法律意见8394篇。界面下半部分是“法务 Q&A”部分，列出了一些法律相关的选项，例如“国际法务接入口人”、“国内法务接入口人”、“国际法律协议合同办理指引”和“国内法律协议合同办理指引”。界面底部有三个导航按钮，分别是“首页”、“模板”和“我的”。
+
+#### 按国家名维度搜索
+法务查询者在国际小超人钉钉小程序的搜索框中进行检索时采用typeahead，只能下拉选择法务中台中有的国家名称。
+![按国家名维度搜索](2.png)
+在这张图像中，用户正在一个名为“法律意见查询”的应用中进行国家名称的搜索。用户在搜索框中输入国家名称时，系统会提供下拉建议。这些建议基于 typeahead 功能，从法务中台中筛选出匹配的国家名称供用户选择。目前，搜索结果包含了“中国”和“菲律宾”两个具体的国家名称，其它显示为“国家名”。用户可以通过下拉菜单快速选择所需的国家名称。
+
+#### 检索结果
+法务查询者可根据国际小超人钉钉小程序UI上的滚筒切换业务线
+![检索结果](3.png)
+这张图片展示了一个移动应用的界面，界面标题为“法律意见详情”。用户可以根据具体情况切换业务线。界面中有多个字段，包括“国家名称”、“国家情况描述”、“业务线”、“产品法规分析”和“签约主体”。第一张截图显示了详细的法律情报信息，包含区域名称、区域情况描述、业务线和产品法规概述等字段。第二张截图显示了“法律意见详情”界面，其中列出了国家名称、国家情况描述、业务线、产品法规分析和签约主体。第三张截图与第二张相似，但显示了选项的可选择状态。最下方有“取消”和“确定”的按钮。
+法务查询者从国家详情中的业务线名列表中选出要查看的业务线。
+
+#### 查看法律意见详情
+国际小超人钉钉小程序用国家代码和业务代码做参数，查询法律意见详情，然后将法律意见详情展示给法务查询者。

tests/data/embedding/2.query.md

@@ -0,0 +1 @@
+业务线UI有哪些操作？

tests/data/embedding/3.answer.md

@@ -0,0 +1,7 @@
+国家/区域导游详情 & 法律意见详情 查询
+Description：根据国家code查询国家/区域导游信息详情
+ID: 8
+HTTP METHOD: GET
+Endpoint: /contract/country/navigate.json
+Input Parameters: |名称|描述|类型(长度)|必选|备注| | :- | :- | :-: | :- | :- | |countryCode|国家code|string|√||
+Returns: |名称|描述|类型(长度)|必选|备注| | :- | :- | :-: | :- | :- | |success|业务处理成功true，否则false|boolean|√|只判断这个属性即可| |message|错误信息，可以用来提示|string|√|| |code|返回状态码|string|√|| |data|国家/区域导游详情|object|√|| |-> country||||| |-> -> id|id|integer|√|| |-> -> country|国家code|string|√|| |-> -> countryName|国家中文名称|string|√|| |-> -> countryNameEn|国家英文名称|string|√|| |-> -> content|国家导游中文详情json数组，具体格式见下示例|list of object|√|| |-> -> -> title|标题|object|√|| |-> -> -> -> title|中文标题|string||| |-> -> -> -> titleEn|英文标题|string||| |-> -> -> contentList|标题下面的文字描述列表|list of object|√|| |-> -> -> -> detail|内容中文详情|string|√|| |-> -> -> -> detailEn|内容英文详情|string|√|| |-> -> -> -> url|超链接|string||| |-> legal|法务信息|object||| |-> -> country|国家code|string|√|| |-> -> businessList|业务线列表|list of object||| |-> -> -> id|id|integer||新增时不传，修改时传递| |-> -> -> business|业务线code|string|√|| |-> -> -> businessName|业务线中文名称|string|√|| |-> -> -> businessNameEn|业务线英文名称|string|√|| |-> -> -> content|业务线json，具体如下|object|√|| |-> -> -> -> detailEn|具体的详情英文内容|string|√|| |-> -> -> -> detail|具体的详情内容|string|√||

tests/data/embedding/3.knowledge.md

@@ -0,0 +1,189 @@
+## Interfaces
+- 用户登录
+  - Description: 用户从小程序/微应用发起请求，需要验证用户的合法身份才能正常处理。
+  - ID: 1
+  - HTTP METHOD: GET
+  - Endpoint: `/sup/login.json`
+  - Input Parameters:
+    |名称|描述|类型(长度)|必选|备注|
+    | :- | :- | :-: | :- | :- |
+    |authCode|用户临时免登授权码|String(64)|√||
+    |loginTypeEnum|登录类型|String(20)|√||
+    |authCorpId|用户所在企业/组织id|String(64)||微应用免登时传递|
+    |app|应用标识|String(3)|√||
+  - Returns:
+    |名称|描述|类型(长度)|必选|备注|
+    | :- | :- | :-: | :- | :- |
+    |success|业务处理成功与否，成功true，否则false|boolean|√|只判断这个属性即可|
+    |message|错误信息，可以用来提示|string|√||
+    |code|返回状态码|string|√||
+    |data|用户的sessionId|string|√||
+- 根据sessionId查询用户详细信息
+  - Description: 查询当前用户的详细信息，如 staffId，unionId，name，avatar等信息
+  - ID: 2
+  - HTTP METHOD: GET
+  - Endpoint: `/sup/user.json`
+  - Input Parameters:
+    |名称|描述|类型(长度)|必选|备注|
+    | :- | :- | :-: | :- | :- |
+    |NDA_SESSION|用户sessionId|String(64)|√||
+  - Returns:
+    |名称|描述|类型(长度)|必选|备注|
+    | :- | :- | :-: | :- | :- |
+    |success|业务处理成功与否，成功true，否则false|boolean|√|只判断这个属性即可|
+    |message|错误信息，可以用来提示|string|√||
+    |code|返回状态码|string|√||
+    |data|用户的详细信息|object|√||
+    |-> corpId|当前用户企业 钉钉ID(小程序端会拿不到该信息)|string|√||
+    |-> corpName|当前用户企业名称(小程序端会拿不到该信息)|string|√||
+    |-> staffId|员工在当前企业内的唯一标识，也称staffId(小程序端会拿不到该信息)|string|√||
+    |-> unionId|员工在当前开发者企业账号范围内的唯一标识，系统生成，固定值，不会改变。|string|√||
+    |-> name|当前用户的名称(小程序端会拿不到该信息)|string|√||
+    |-> avatar|头像图片URL|string|√||
+- 查询国家情况描述
+  - Description: 根据国家code查询国家情况描述
+  - ID: 3
+  - HTTP METHOD: GET
+  - Endpoint: `/sup/country/detail.json`
+  - Input Parameters:
+    |名称|描述|类型(长度)|必选|备注|
+    | :- | :- | :-: | :- | :- |
+    |countryCode|国家code|string|√||
+  - Returns:
+    |名称|描述|类型(长度)|必选|备注|
+    | :- | :- | :-: | :- | :- |
+    |success|业务处理成功true，否则false|boolean|√|只判断这个属性即可|
+    |message|错误信息，可以用来提示|string|√||
+    |code|返回状态码|string|√||
+    |data|国家情况描述|object|√||
+    |-> id|id|integer|√||
+    |-> countryName|国家名称|string|√||
+    |-> countryCode|国家code|string|√||
+    |-> detail|产品法规分析|string|√||
+- 查询产品法规分析（法律意见详情）
+  - Description: 根据国家和业务线查询产品法规分析
+  - ID: 4
+  - HTTP METHOD: GET
+  - Endpoint: `/sup/legal/detail.json`
+  - Input Parameters:
+    |名称|描述|类型(长度)|必选|备注|
+    | :- | :- | :-: | :- | :- |
+    |countryCode|国家code|string|√||
+    |businessCode|业务线code|string|√||
+  - Returns:
+    |名称|描述|类型(长度)|必选|备注|
+    | :- | :- | :-: | :- | :- |
+    |success|业务处理成功true，否则false|boolean|√|只判断这个属性即可|
+    |message|错误信息，可以用来提示|string|√||
+    |code|返回状态码|string|√||
+    |data|法律意见详情|object|√||
+    |-> id|id|integer|√||
+    |-> countryName|国家名称|string|√||
+    |-> countryCode|国家code|string|√||
+    |-> businessLine|业务线|string|√||
+    |-> businessCode|业务线code|string|√||
+    |-> detail|产品法规分析|string|√||
+    |-> signEntity|签约主体|string|√||
+- 查询法律意见总数
+  - Description: 法律意见总数查询
+  - ID: 5
+  - HTTP METHOD: GET
+  - Endpoint: `/sup/legal/count.json`
+  - Input Parameters:
+    |名称|描述|类型(长度)|必选|备注|
+    | :- | :- | :-: | :- | :- |
+  - Returns:
+    |名称|描述|类型(长度)|必选|备注|
+    | :- | :- | :-: | :- | :- |
+    |success|业务处理成功true，否则false|boolean|√|只判断这个属性即可|
+    |message|错误信息，可以用来提示|string|√||
+    |code|返回状态码|string|√||
+    |data|总数|integer|√||
+- 查询所有国家和业务线信息列表
+  - Description: 查询所有国家和业务线信息列表
+  - ID: 6
+  - HTTP METHOD: GET
+  - Endpoint: `/sup/legal/country/list.json`
+  - Input Parameters:
+    |名称|描述|类型(长度)|必选|备注|
+    | :- | :- | :-: | :- | :- |
+  - Returns:
+    |名称|描述|类型(长度)|必选|备注|
+    | :- | :- | :-: | :- | :- |
+    |success|业务处理成功true，否则false|boolean|√|只判断这个属性即可|
+    |message|错误信息，可以用来提示|string|√||
+    |code|返回状态码|string|√||
+    |data|所有数据列表|list of object|√||
+    |-> country|国家code|string|√||
+    |-> business|业务线code|string|√||
+    |-> dataType|数据类型|string|√||
+    |-> businessName|业务线名|string|√||
+    |-> countryName|国家名|string|√||
+    |-> businessNameEn|业务线名(英文)|string|√||
+- 调用法务中台antlaw接口
+  - ID: 7
+- 国家/区域导游详情 & 法律意见详情 查询
+  - Description：根据国家code查询国家/区域导游信息详情
+  - ID: 8
+  - HTTP METHOD: GET
+  - Endpoint: `/contract/country/navigate.json`
+  - Input Parameters:
+    |名称|描述|类型(长度)|必选|备注|
+    | :- | :- | :-: | :- | :- |
+    |countryCode|国家code|string|√||
+  - Returns:
+    |名称|描述|类型(长度)|必选|备注|
+    | :- | :- | :-: | :- | :- |
+    |success|业务处理成功true，否则false|boolean|√|只判断这个属性即可|
+    |message|错误信息，可以用来提示|string|√||
+    |code|返回状态码|string|√||
+    |data|国家/区域导游详情|object|√||
+    |-> country|||||
+    |-> -> id|id|integer|√||
+    |-> -> country|国家code|string|√||
+    |-> -> countryName|国家中文名称|string|√||
+    |-> -> countryNameEn|国家英文名称|string|√||
+    |-> -> content|国家导游中文详情json数组，具体格式见下示例|list of object|√||
+    |-> -> -> title|标题|object|√||
+    |-> -> -> -> title|中文标题|string|||
+    |-> -> -> -> titleEn|英文标题|string|||
+    |-> -> -> contentList|标题下面的文字描述列表|list of object|√||
+    |-> -> -> -> detail|内容中文详情|string|√||
+    |-> -> -> -> detailEn|内容英文详情|string|√||
+    |-> -> -> -> url|超链接|string|||
+    |-> legal|法务信息|object|||
+    |-> -> country|国家code|string|√||
+    |-> -> businessList|业务线列表|list of object|||
+    |-> -> -> id|id|integer||新增时不传，修改时传递|
+    |-> -> -> business|业务线code|string|√||
+    |-> -> -> businessName|业务线中文名称|string|√||
+    |-> -> -> businessNameEn|业务线英文名称|string|√||
+    |-> -> -> content|业务线json，具体如下|object|√||
+    |-> -> -> -> detailEn|具体的详情英文内容|string|√||
+    |-> -> -> -> detail|具体的详情内容|string|√||
+- 国家/区域导游列表分页查询
+  - Description: 分页查询国家/区域列表
+  - ID: 9
+  - HTTP METHOD: GET
+  - Endpoint: `/contract/country/list.json`
+  - Input Parameters:
+    |名称|描述|类型(长度)|必选|备注|
+    | :- | :- | :-: | :- | :- |
+    |pageSize|分页大小|integer|√|>=1|
+    |pageNum|分页大小|integer|√|>=1|
+    |country|国家code|string|||
+    |business|业务线code|string|||
+  - Returns:
+    |名称|描述|类型(长度)|必选|备注|
+    | :- | :- | :-: | :- | :- |
+    |success|业务处理成功true，否则false|boolean|√|只判断这个属性即可|
+    |message|错误信息，可以用来提示|string|√||
+    |code|返回状态码|string|√||
+    |data|国家/区域导游详情|list of object|√||
+    |-> id|id|integer|√||
+    |-> country|国家code|string|√||
+    |-> countryName|国家中文名称|string|√||
+    |-> countryNameEn|国家英文名称|string|√||
+    |-> gmtCreate|创建时间|string|√||
+    |-> gmtModified|更新时间|string|√||
+    |total|数据总量|integer|√||

tests/data/embedding/3.query.md

@@ -0,0 +1 @@
+根据国家code查询国家业务线列表

tests/metagpt/rag/test_large_pdf.py

@@ -0,0 +1,55 @@
+import pytest
+
+from metagpt.config2 import Config
+from metagpt.const import TEST_DATA_PATH
+from metagpt.rag.engines import SimpleEngine
+from metagpt.rag.factories.embedding import RAGEmbeddingFactory
+from metagpt.utils.common import aread
+
+
+@pytest.mark.skip
+@pytest.mark.parametrize(
+    ("knowledge_filename", "query_filename", "answer_filename"),
+    [
+        (
+            TEST_DATA_PATH / "embedding/2.knowledge.md",
+            TEST_DATA_PATH / "embedding/2.query.md",
+            TEST_DATA_PATH / "embedding/2.answer.md",
+        ),
+        (
+            TEST_DATA_PATH / "embedding/3.knowledge.md",
+            TEST_DATA_PATH / "embedding/3.query.md",
+            TEST_DATA_PATH / "embedding/3.answer.md",
+        ),
+    ],
+)
+@pytest.mark.asyncio
+async def test_large_pdf(knowledge_filename, query_filename, answer_filename):
+    Config.default(reload=True)  # `config.embedding.model = "text-embedding-ada-002"` changes the cache.
+
+    engine = SimpleEngine.from_docs(
+        input_files=[knowledge_filename],
+    )
+
+    query = await aread(filename=query_filename)
+    rsp = await engine.aretrieve(query)
+    assert rsp
+
+    config = Config.default()
+    config.embedding.model = "text-embedding-ada-002"
+    factory = RAGEmbeddingFactory(config)
+    embedding = factory.get_rag_embedding()
+    answer = await aread(filename=answer_filename)
+    answer_embedding = await embedding.aget_text_embedding(answer)
+    similarity = 0
+    for i in rsp:
+        rsp_embedding = await embedding.aget_query_embedding(i.text)
+        v = embedding.similarity(answer_embedding, rsp_embedding)
+        similarity = max(similarity, v)
+
+    print(similarity)
+    assert similarity > 0.9
+
+
+if __name__ == "__main__":
+    pytest.main([__file__, "-s"])

tests/metagpt/tools/libs/test_editor.py

@@ -1,7 +1,7 @@
 import pytest
 
 from metagpt.const import TEST_DATA_PATH
-from metagpt.tools.libs.editor import Editor, FileBlock
+from metagpt.tools.libs.editor import Editor
 
 TEST_FILE_CONTENT = """
 # this is line one
@@ -13,31 +13,24 @@ def test_function_for_fm():
     # this is the 7th line
 """.strip()
 
-TEST_FILE_PATH = TEST_DATA_PATH / "tools/test_script_for_editor.py"
+WINDOW = 100
 
 
 @pytest.fixture
-def test_file():
-    with open(TEST_FILE_PATH, "w") as f:
-        f.write(TEST_FILE_CONTENT)
-    yield
-    with open(TEST_FILE_PATH, "w") as f:
-        f.write("")
+def temp_file_path(tmp_path):
+    assert tmp_path is not None
+    temp_file_path = tmp_path / "a.txt"
+    yield temp_file_path
+    temp_file_path.unlink()
 
 
-EXPECTED_SEARCHED_BLOCK = FileBlock(
-    file_path=str(TEST_FILE_PATH),
-    block_content='001|# this is line one\n002|def test_function_for_fm():\n003|    "some docstring"\n004|    a = 1\n005|    b = 2\n',
-    block_start_line=1,
-    block_end_line=5,
-    symbol="def test_function_for_fm",
-    symbol_line=2,
-)
-
-
-def test_search_content(test_file):
-    block = Editor().search_content("def test_function_for_fm", root_path=TEST_DATA_PATH, window=3)
-    assert block == EXPECTED_SEARCHED_BLOCK
+@pytest.fixture
+def temp_py_file(tmp_path):
+    assert tmp_path is not None
+    temp_file_path = tmp_path / "test_script_for_editor.py"
+    temp_file_path.write_text(TEST_FILE_CONTENT)
+    yield temp_file_path
+    temp_file_path.unlink()
 
 
 EXPECTED_CONTENT_AFTER_REPLACE = """
@@ -50,32 +43,43 @@ def test_function_for_fm():
 """.strip()
 
 
-@pytest.mark.skip
-def test_replace_content(test_file):
-    Editor().write_content(
-        file_path=str(TEST_FILE_PATH),
-        start_line=3,
-        end_line=5,
-        new_block_content="    # This is the new line A replacing lines 3 to 5.\n    # This is the new line B.",
+def test_replace_content(temp_py_file):
+    editor = Editor()
+    editor._edit_file_impl(
+        file_name=temp_py_file,
+        start=3,
+        end=5,
+        content="    # This is the new line A replacing lines 3 to 5.\n    # This is the new line B.",
+        is_insert=False,
+        is_append=False,
     )
-    with open(TEST_FILE_PATH, "r") as f:
+    with open(temp_py_file, "r") as f:
         new_content = f.read()
-    assert new_content == EXPECTED_CONTENT_AFTER_REPLACE
+    assert new_content.strip() == EXPECTED_CONTENT_AFTER_REPLACE.strip()
 
 
 EXPECTED_CONTENT_AFTER_DELETE = """
 # this is line one
 def test_function_for_fm():
+
     c = 3
     # this is the 7th line
 """.strip()
 
 
-def test_delete_content(test_file):
-    Editor().write_content(file_path=str(TEST_FILE_PATH), start_line=3, end_line=5)
-    with open(TEST_FILE_PATH, "r") as f:
+def test_delete_content(temp_py_file):
+    editor = Editor()
+    editor._edit_file_impl(
+        file_name=temp_py_file,
+        start=3,
+        end=5,
+        content="",
+        is_insert=False,
+        is_append=False,
+    )
+    with open(temp_py_file, "r") as f:
         new_content = f.read()
-    assert new_content == EXPECTED_CONTENT_AFTER_DELETE
+    assert new_content.strip() == EXPECTED_CONTENT_AFTER_DELETE.strip()
 
 
 EXPECTED_CONTENT_AFTER_INSERT = """
@@ -90,39 +94,16 @@ def test_function_for_fm():
 """.strip()
 
 
-@pytest.mark.skip
-def test_insert_content(test_file):
-    Editor().write_content(
-        file_path=str(TEST_FILE_PATH),
-        start_line=3,
-        end_line=-1,
-        new_block_content="    # This is the new line to be inserted, at line 3",
+def test_insert_content(temp_py_file):
+    editor = Editor(enable_auto_lint=True)
+    editor.insert_content_at_line(
+        file_name=temp_py_file,
+        line_number=3,
+        content="    # This is the new line to be inserted, at line 3",
     )
-    with open(TEST_FILE_PATH, "r") as f:
+    with open(temp_py_file, "r") as f:
         new_content = f.read()
-    assert new_content == EXPECTED_CONTENT_AFTER_INSERT
-
-
-@pytest.mark.skip
-def test_new_content_wrong_indentation(test_file):
-    msg = Editor().write_content(
-        file_path=str(TEST_FILE_PATH),
-        start_line=3,
-        end_line=-1,
-        new_block_content="    This is the new line to be inserted, at line 3",  # omit # should throw a syntax error
-    )
-    assert "failed" in msg
-
-
-@pytest.mark.skip
-def test_new_content_format_issue(test_file):
-    msg = Editor().write_content(
-        file_path=str(TEST_FILE_PATH),
-        start_line=3,
-        end_line=-1,
-        new_block_content="    # This is the new line to be inserted, at line 3  ",  # trailing spaces are format issue only, and should not throw an error
-    )
-    assert "failed" not in msg
+    assert new_content.strip() == EXPECTED_CONTENT_AFTER_INSERT.strip()
 
 
 @pytest.mark.parametrize(
@@ -151,5 +132,518 @@ async def test_read_files(filename):
         assert file_block.block_content
 
 
+def _numbered_test_lines(start, end) -> str:
+    return ("\n".join(f"{i}|" for i in range(start, end + 1))) + "\n"
+
+
+def _generate_test_file_with_lines(temp_path, num_lines) -> str:
+    file_path = temp_path / "test_file.py"
+    file_path.write_text("\n" * num_lines)
+    return file_path
+
+
+def _generate_ruby_test_file_with_lines(temp_path, num_lines) -> str:
+    file_path = temp_path / "test_file.rb"
+    file_path.write_text("\n" * num_lines)
+    return file_path
+
+
+def _calculate_window_bounds(current_line, total_lines, window_size):
+    half_window = window_size // 2
+    if current_line - half_window < 0:
+        start = 1
+        end = window_size
+    else:
+        start = current_line - half_window
+        end = current_line + half_window
+    return start, end
+
+
+def test_open_file_unexist_path():
+    editor = Editor()
+    with pytest.raises(FileNotFoundError):
+        editor.open_file("/unexist/path/a.txt")
+
+
+def test_open_file(temp_file_path):
+    editor = Editor()
+    temp_file_path.write_text("Line 1\nLine 2\nLine 3\nLine 4\nLine 5")
+
+    result = editor.open_file(str(temp_file_path))
+
+    assert result is not None
+    expected = (
+        f"[File: {temp_file_path} (5 lines total)]\n"
+        "(this is the beginning of the file)\n"
+        "1|Line 1\n"
+        "2|Line 2\n"
+        "3|Line 3\n"
+        "4|Line 4\n"
+        "5|Line 5\n"
+        "(this is the end of the file)"
+    )
+    assert result.split("\n") == expected.split("\n")
+
+
+def test_open_file_with_indentation(temp_file_path):
+    editor = Editor()
+    temp_file_path.write_text("Line 1\n    Line 2\nLine 3\nLine 4\nLine 5")
+
+    result = editor.open_file(str(temp_file_path))
+    assert result is not None
+    expected = (
+        f"[File: {temp_file_path} (5 lines total)]\n"
+        "(this is the beginning of the file)\n"
+        "1|Line 1\n"
+        "2|    Line 2\n"
+        "3|Line 3\n"
+        "4|Line 4\n"
+        "5|Line 5\n"
+        "(this is the end of the file)"
+    )
+    assert result.split("\n") == expected.split("\n")
+
+
+def test_open_file_long(temp_file_path):
+    editor = Editor()
+    content = "\n".join([f"Line {i}" for i in range(1, 1001)])
+    temp_file_path.write_text(content)
+
+    result = editor.open_file(str(temp_file_path), 1, 50)
+    assert result is not None
+    expected = f"[File: {temp_file_path} (1000 lines total)]\n"
+    expected += "(this is the beginning of the file)\n"
+    for i in range(1, 51):
+        expected += f"{i}|Line {i}\n"
+    expected += "(950 more lines below)"
+    assert result.split("\n") == expected.split("\n")
+
+
+def test_open_file_long_with_lineno(temp_file_path):
+    editor = Editor()
+    content = "\n".join([f"Line {i}" for i in range(1, 1001)])
+    temp_file_path.write_text(content)
+
+    cur_line = 100
+
+    result = editor.open_file(str(temp_file_path), cur_line)
+    assert result is not None
+    expected = f"[File: {temp_file_path} (1000 lines total)]\n"
+    start, end = _calculate_window_bounds(cur_line, 1000, WINDOW)
+    if start == 1:
+        expected += "(this is the beginning of the file)\n"
+    else:
+        expected += f"({start - 1} more lines above)\n"
+    for i in range(start, end + 1):
+        expected += f"{i}|Line {i}\n"
+    if end == 1000:
+        expected += "(this is the end of the file)\n"
+    else:
+        expected += f"({1000 - end} more lines below)"
+    assert result.split("\n") == expected.split("\n")
+
+
+def test_create_file_unexist_path():
+    editor = Editor()
+    with pytest.raises(FileNotFoundError):
+        editor.create_file("/unexist/path/a.txt")
+
+
+def test_create_file(temp_file_path):
+    editor = Editor()
+    result = editor.create_file(str(temp_file_path))
+
+    expected = f"[File {temp_file_path} created.]"
+    assert result.split("\n") == expected.split("\n")
+
+
+def test_goto_line(temp_file_path):
+    editor = Editor()
+    total_lines = 1000
+    content = "\n".join([f"Line {i}" for i in range(1, total_lines + 1)])
+    temp_file_path.write_text(content)
+
+    result = editor.open_file(str(temp_file_path))
+    assert result is not None
+
+    expected = f"[File: {temp_file_path} ({total_lines} lines total)]\n"
+    expected += "(this is the beginning of the file)\n"
+    for i in range(1, WINDOW + 1):
+        expected += f"{i}|Line {i}\n"
+    expected += f"({total_lines - WINDOW} more lines below)"
+    assert result.split("\n") == expected.split("\n")
+
+    result = editor.goto_line(500)
+
+    assert result is not None
+
+    cur_line = 500
+    expected = f"[File: {temp_file_path} ({total_lines} lines total)]\n"
+    start, end = _calculate_window_bounds(cur_line, total_lines, WINDOW)
+    if start == 1:
+        expected += "(this is the beginning of the file)\n"
+    else:
+        expected += f"({start - 1} more lines above)\n"
+    for i in range(start, end + 1):
+        expected += f"{i}|Line {i}\n"
+    if end == total_lines:
+        expected += "(this is the end of the file)\n"
+    else:
+        expected += f"({total_lines - end} more lines below)"
+    assert result.split("\n") == expected.split("\n")
+
+
+def test_goto_line_negative(temp_file_path):
+    editor = Editor()
+    content = "\n".join([f"Line {i}" for i in range(1, 5)])
+    temp_file_path.write_text(content)
+
+    editor.open_file(str(temp_file_path))
+    with pytest.raises(ValueError):
+        editor.goto_line(-1)
+
+
+def test_goto_line_out_of_bound(temp_file_path):
+    editor = Editor()
+    content = "\n".join([f"Line {i}" for i in range(1, 5)])
+    temp_file_path.write_text(content)
+
+    editor.open_file(str(temp_file_path))
+    with pytest.raises(ValueError):
+        editor.goto_line(100)
+
+
+def test_scroll_down(temp_file_path):
+    editor = Editor()
+    total_lines = 1000
+    content = "\n".join([f"Line {i}" for i in range(1, total_lines + 1)])
+    temp_file_path.write_text(content)
+    result = editor.open_file(str(temp_file_path))
+    assert result is not None
+
+    expected = f"[File: {temp_file_path} ({total_lines} lines total)]\n"
+    start, end = _calculate_window_bounds(1, total_lines, WINDOW)
+    if start == 1:
+        expected += "(this is the beginning of the file)\n"
+    else:
+        expected += f"({start - 1} more lines above)\n"
+    for i in range(start, end + 1):
+        expected += f"{i}|Line {i}\n"
+    if end == total_lines:
+        expected += "(this is the end of the file)"
+    else:
+        expected += f"({total_lines - end} more lines below)"
+    assert result.split("\n") == expected.split("\n")
+
+    result = editor.scroll_down()
+
+    assert result is not None
+
+    expected = f"[File: {temp_file_path} ({total_lines} lines total)]\n"
+    start, end = _calculate_window_bounds(WINDOW + 1, total_lines, WINDOW)
+    if start == 1:
+        expected += "(this is the beginning of the file)\n"
+    else:
+        expected += f"({start - 1} more lines above)\n"
+    for i in range(start, end + 1):
+        expected += f"{i}|Line {i}\n"
+    if end == total_lines:
+        expected += "(this is the end of the file)\n"
+    else:
+        expected += f"({total_lines - end} more lines below)"
+    assert result.split("\n") == expected.split("\n")
+
+
+def test_scroll_up(temp_file_path):
+    editor = Editor()
+    total_lines = 1000
+    content = "\n".join([f"Line {i}" for i in range(1, total_lines + 1)])
+    temp_file_path.write_text(content)
+
+    cur_line = 300
+
+    result = editor.open_file(str(temp_file_path), cur_line)
+    assert result is not None
+
+    expected = f"[File: {temp_file_path} ({total_lines} lines total)]\n"
+    start, end = _calculate_window_bounds(cur_line, total_lines, WINDOW)
+    if start == 1:
+        expected += "(this is the beginning of the file)\n"
+    else:
+        expected += f"({start - 1} more lines above)\n"
+    for i in range(start, end + 1):
+        expected += f"{i}|Line {i}\n"
+    if end == total_lines:
+        expected += "(this is the end of the file)\n"
+    else:
+        expected += f"({total_lines - end} more lines below)"
+    assert result.split("\n") == expected.split("\n")
+    result = editor.scroll_up()
+    assert result is not None
+
+    cur_line = cur_line - WINDOW
+
+    expected = f"[File: {temp_file_path} ({total_lines} lines total)]\n"
+    start, end = _calculate_window_bounds(cur_line, total_lines, WINDOW)
+    if start == 1:
+        expected += "(this is the beginning of the file)\n"
+    else:
+        expected += f"({start - 1} more lines above)\n"
+    for i in range(start, end + 1):
+        expected += f"{i}|Line {i}\n"
+    if end == total_lines:
+        expected += "(this is the end of the file)\n"
+    else:
+        expected += f"({total_lines - end} more lines below)"
+    assert result.split("\n") == expected.split("\n")
+
+
+def test_scroll_down_edge(temp_file_path):
+    editor = Editor()
+    content = "\n".join([f"Line {i}" for i in range(1, 10)])
+    temp_file_path.write_text(content)
+
+    result = editor.open_file(str(temp_file_path))
+    assert result is not None
+
+    expected = f"[File: {temp_file_path} (9 lines total)]\n"
+    expected += "(this is the beginning of the file)\n"
+    for i in range(1, 10):
+        expected += f"{i}|Line {i}\n"
+    expected += "(this is the end of the file)"
+
+    result = editor.scroll_down()
+    assert result is not None
+
+    assert result.split("\n") == expected.split("\n")
+
+
+def test_print_window_internal(temp_file_path):
+    editor = Editor()
+    editor.create_file(str(temp_file_path))
+    with open(temp_file_path, "w") as file:
+        for i in range(1, 101):
+            file.write(f"Line `{i}`\n")
+
+    current_line = 50
+    window = 2
+
+    result = editor._print_window(temp_file_path, current_line, window)
+    expected = "(48 more lines above)\n" "49|Line `49`\n" "50|Line `50`\n" "51|Line `51`\n" "(49 more lines below)"
+    assert result == expected
+
+
+def test_open_file_large_line_number(temp_file_path):
+    editor = Editor()
+    editor.create_file(str(temp_file_path))
+    with open(temp_file_path, "w") as file:
+        for i in range(1, 1000):
+            file.write(f"Line `{i}`\n")
+
+    current_line = 800
+    window = 100
+
+    result = editor.open_file(str(temp_file_path), current_line, window)
+
+    expected = f"[File: {temp_file_path} (999 lines total)]\n"
+    expected += "(749 more lines above)\n"
+    for i in range(750, 850 + 1):
+        expected += f"{i}|Line `{i}`\n"
+    expected += "(149 more lines below)"
+    assert result == expected
+
+
+def test_open_file_large_line_number_consecutive_diff_window(temp_file_path):
+    editor = Editor()
+    editor.create_file(str(temp_file_path))
+    total_lines = 1000
+    with open(temp_file_path, "w") as file:
+        for i in range(1, total_lines + 1):
+            file.write(f"Line `{i}`\n")
+
+    current_line = 800
+    cur_window = 300
+
+    result = editor.open_file(str(temp_file_path), current_line, cur_window)
+
+    expected = f"[File: {temp_file_path} ({total_lines} lines total)]\n"
+    start, end = _calculate_window_bounds(current_line, total_lines, cur_window)
+    if start == 1:
+        expected += "(this is the beginning of the file)\n"
+    else:
+        expected += f"({start - 1} more lines above)\n"
+    for i in range(current_line - cur_window // 2, current_line + cur_window // 2 + 1):
+        expected += f"{i}|Line `{i}`\n"
+    if end == total_lines:
+        expected += "(this is the end of the file)\n"
+    else:
+        expected += f"({total_lines - end} more lines below)"
+    assert result == expected
+
+    current_line = current_line - WINDOW
+
+    result = editor.scroll_up()
+
+    expected = f"[File: {temp_file_path} ({total_lines} lines total)]\n"
+    start, end = _calculate_window_bounds(current_line, total_lines, WINDOW)
+    if start == 1:
+        expected += "(this is the beginning of the file)\n"
+    else:
+        expected += f"({start - 1} more lines above)\n"
+    for i in range(start, end + 1):
+        expected += f"{i}|Line `{i}`\n"
+    if end == total_lines:
+        expected += "(this is the end of the file)\n"
+    else:
+        expected += f"({total_lines - end} more lines below)"
+    assert result.split("\n") == expected.split("\n")
+
+
+EXPECTED_CONTENT_AFTER_REPLACE_TEXT = """
+# this is line one
+def test_function_for_fm():
+    "some docstring"
+    a = 1
+    b = 9
+    c = 3
+    # this is the 7th line
+""".strip()
+
+
+def test_edit_file_by_replace(temp_py_file):
+    editor = Editor()
+    editor.edit_file_by_replace(file_name=str(temp_py_file), to_replace="    b = 2", new_content="    b = 9")
+    with open(temp_py_file, "r") as f:
+        new_content = f.read()
+    assert new_content.strip() == EXPECTED_CONTENT_AFTER_REPLACE_TEXT.strip()
+
+
+def test_append_file(temp_file_path):
+    editor = Editor()
+    # 写入初始内容
+    initial_content = "Line 1\nLine 2\nLine 3\n"
+    temp_file_path.write_text(initial_content)
+
+    # 追加内容到文件
+    append_content = "Line 4\nLine 5\n"
+
+    result = editor.append_file(str(temp_file_path), append_content)
+
+    # 预期内容
+    expected_content = initial_content + append_content
+
+    # 读取文件并断言内容与预期一致
+    with open(temp_file_path, "r") as f:
+        new_content = f.read()
+    assert new_content == expected_content
+
+    # 输出的预期结果
+    expected_output = (
+        f"[File: {temp_file_path.resolve()} (5 lines total after edit)]\n"
+        "(this is the beginning of the file)\n"
+        "1|Line 1\n"
+        "2|Line 2\n"
+        "3|Line 3\n"
+        "4|Line 4\n"
+        "5|Line 5\n"
+        "(this is the end of the file)\n"
+        "[File updated (edited at line 3). Please review the changes and make sure they are correct (correct indentation, no duplicate lines, etc). Edit the file again if necessary.]"
+    )
+
+    assert result.split("\n") == expected_output.split("\n")
+
+
+def test_search_dir(tmp_path):
+    editor = Editor()
+    dir_path = tmp_path / "test_dir"
+    dir_path.mkdir()
+
+    # Create some files with specific content
+    (dir_path / "file1.txt").write_text("This is a test file with some content.")
+    (dir_path / "file2.txt").write_text("Another file with different content.")
+    sub_dir = dir_path / "sub_dir"
+    sub_dir.mkdir()
+    (sub_dir / "file3.txt").write_text("This file is inside a sub directory with some content.")
+
+    search_term = "some content"
+
+    result = editor.search_dir(search_term, str(dir_path))
+
+    assert "file1.txt" in result
+    assert "file3.txt" in result
+    assert "Another file with different content." not in result
+
+
+def test_search_file(temp_file_path):
+    editor = Editor()
+    file_path = temp_file_path
+    file_path.write_text("This is a test file with some content.\nAnother line with more content.")
+
+    search_term = "some content"
+
+    result = editor.search_file(search_term, str(file_path))
+
+    assert "Line 1: This is a test file with some content." in result
+    assert "Line 2: Another line with more content." not in result
+
+
+def test_find_file(tmp_path):
+    editor = Editor()
+    dir_path = tmp_path / "test_dir"
+    dir_path.mkdir()
+
+    # Create some files with specific names
+    (dir_path / "file1.txt").write_text("Content of file 1.")
+    (dir_path / "file2.txt").write_text("Content of file 2.")
+    sub_dir = dir_path / "sub_dir"
+    sub_dir.mkdir()
+    (sub_dir / "file3.txt").write_text("Content of file 3.")
+
+    file_name = "file1.txt"
+
+    result = editor.find_file(file_name, str(dir_path))
+
+    assert "file1.txt" in result
+    assert "file2.txt" not in result
+    assert "file3.txt" not in result
+
+
+# Test data for _append_impl method
+TEST_LINES = ["First line\n", "Second line\n", "Third line\n"]
+
+NEW_CONTENT = "Appended line\n"
+
+EXPECTED_APPEND_NON_EMPTY_FILE = ["First line\n", "Second line\n", "Third line\n", "Appended line\n"]
+
+EXPECTED_APPEND_EMPTY_FILE = ["Appended line\n"]
+
+
+def test_append_non_empty_file():
+    editor = Editor()
+    lines = TEST_LINES.copy()
+    content, n_added_lines = editor._append_impl(lines, NEW_CONTENT)
+
+    assert content.splitlines(keepends=True) == EXPECTED_APPEND_NON_EMPTY_FILE
+    assert n_added_lines == 1
+
+
+def test_append_empty_file():
+    editor = Editor()
+    lines = []
+    content, n_added_lines = editor._append_impl(lines, NEW_CONTENT)
+
+    assert content.splitlines(keepends=True) == EXPECTED_APPEND_EMPTY_FILE
+    assert n_added_lines == 1
+
+
+def test_append_to_single_empty_line_file():
+    editor = Editor()
+    lines = [""]
+    content, n_added_lines = editor._append_impl(lines, NEW_CONTENT)
+
+    assert content.splitlines(keepends=True) == EXPECTED_APPEND_EMPTY_FILE
+    assert n_added_lines == 1
+
+
 if __name__ == "__main__":
     pytest.main([__file__, "-s"])

tests/metagpt/tools/libs/test_index_repo.py

@@ -0,0 +1,32 @@
+import shutil
+
+import pytest
+
+from metagpt.const import DEFAULT_WORKSPACE_ROOT, TEST_DATA_PATH
+from metagpt.tools.libs.index_repo import IndexRepo
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize(("path", "query"), [(TEST_DATA_PATH / "requirements", "业务线")])
+async def test_index_repo(path, query):
+    index_path = DEFAULT_WORKSPACE_ROOT / ".index"
+    repo = IndexRepo(persist_path=str(index_path), root_path=str(path), min_token_count=0)
+    await repo.add([path])
+    await repo.add([path])
+    assert index_path.exists()
+
+    rsp = await repo.search(query)
+    assert rsp
+
+    repo2 = IndexRepo(persist_path=str(index_path), root_path=str(path), min_token_count=0)
+    rsp2 = await repo2.search(query)
+    assert rsp2
+
+    merged_rsp = await repo.merge(query=query, indices_list=[rsp, rsp2])
+    assert merged_rsp
+
+    shutil.rmtree(index_path)
+
+
+if __name__ == "__main__":
+    pytest.main([__file__, "-s"])