🌟 The Multi-Agent Framework: First AI Software Company, Towards Natural Language Programming
feat: Editor.read + large file check
莘权 马aa5b78a37091
diff --git a/metagpt/tools/libs/editor.py b/metagpt/tools/libs/editor.pyindex c5c828ee2..a37c01421 100644--- a/metagpt/tools/libs/editor.py+++ b/metagpt/tools/libs/editor.py@@ -10,10 +10,11 @@ import tempfilefrom pathlib import Pathfrom typing import List, Optional, Union+import tiktokenfrom pydantic import BaseModel, ConfigDictfrom metagpt.const import DEFAULT_WORKSPACE_ROOT-from metagpt.tools.libs.index_repo import IndexRepo+from metagpt.tools.libs.index_repo import DEFAULT_MIN_TOKEN_COUNT, IndexRepofrom metagpt.tools.libs.linter import Linterfrom metagpt.tools.tool_registry import register_toolfrom metagpt.utils.file import File@@ -106,6 +107,11 @@ class Editor(BaseModel):content = await File.read_text_file(path)if not content:return FileBlock(file_path=str(path), block_content="")+ if self.is_large_file(content=content):+ return FileBlock(+ file_path=str(path),+ block_content="The file is too large to read. Use `Editor.similarity_search` to read the file instead.",+ )self.resource.report(str(path), "path")lines = content.splitlines(keepends=True)@@ -959,3 +965,10 @@ class Editor(BaseModel):>>> print(texts)"""return await IndexRepo.cross_repo_search(query=query, file_or_path=file_or_path)++ @staticmethod+ def is_large_file(content: str, mix_token_count: int = 0) -> bool:+ encoding = tiktoken.get_encoding("cl100k_base")+ token_count = len(encoding.encode(content))+ mix_token_count = mix_token_count or DEFAULT_MIN_TOKEN_COUNT+ return token_count >= mix_token_countdiff --git a/tests/metagpt/tools/libs/test_editor.py b/tests/metagpt/tools/libs/test_editor.pyindex 9a6e4b62f..40b6d1426 100644--- a/tests/metagpt/tools/libs/test_editor.py+++ b/tests/metagpt/tools/libs/test_editor.py@@ -737,5 +737,13 @@ async def test_similarity_search(query, filename):save_to.unlink(missing_ok=True)+async def test_read():+ editor = Editor()+ content = await editor.read(str(TEST_DATA_PATH / "pdf/9112674.pdf"))+ assert "similarity_search" in content.block_content++if __name__ == "__main__":pytest.main([__file__, "-s"])