feat: +git tools

metagpt/config2.py

@@ -47,7 +47,7 @@ class Config(CLIParams, YamlModel):
     # Key Parameters
     llm: LLMConfig
 
-    # Global Proxy. Will be used if llm.proxy is not set
+    # Global Proxy. Not used by LLM, but by other tools such as browsers.
     proxy: str = ""
 
     # Tool Parameters

metagpt/tools/libs/git.py

@@ -0,0 +1,65 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+from __future__ import annotations
+
+from pathlib import Path
+
+from metagpt.tools.tool_registry import register_tool
+from metagpt.utils.git_repository import GitRepository
+
+
+@register_tool(tags=["git"])
+async def git_clone(url: str, output_dir: str | Path = None) -> Path:
+    """
+    Clones a Git repository from the given URL.
+
+    Args:
+        url (str): The URL of the Git repository to clone.
+        output_dir (str or Path, optional): The directory where the repository will be cloned.
+            If not provided, the repository will be cloned into the current working directory.
+
+    Returns:
+        Path: The path to the cloned repository.
+
+    Raises:
+        ValueError: If the specified Git root is invalid.
+
+    Example:
+        >>> # git clone to /TO/PATH
+        >>> url = 'https://github.com/geekan/MetaGPT.git'
+        >>> output_dir = "/TO/PATH"
+        >>> repo_dir = await git_clone(url=url, output_dir=output_dir)
+        >>> print(repo_dir)
+        /TO/PATH/MetaGPT
+
+        >>> # git clone to default directory.
+        >>> url = 'https://github.com/geekan/MetaGPT.git'
+        >>> repo_dir = await git_clone(url)
+        >>> print(repo_dir)
+        /WORK_SPACE/downloads/MetaGPT
+    """
+    repo = await GitRepository.clone_from(url, output_dir)
+    return repo.workdir
+
+
+async def git_checkout(repo_dir: str | Path, commit_id: str):
+    """
+    Checks out a specific commit in a Git repository.
+
+    Args:
+        repo_dir (str or Path): The directory containing the Git repository.
+        commit_id (str): The ID of the commit to check out.
+
+    Raises:
+        ValueError: If the specified Git root is invalid.
+
+    Example:
+        >>> repo_dir = '/TO/GIT/REPO'
+        >>> commit_id = 'main'
+        >>> await git_checkout(repo_dir=repo_dir, commit_id=commit_id)
+        git checkout main
+    """
+    repo = GitRepository(local_path=repo_dir, auto_init=False)
+    if not repo.is_valid:
+        ValueError(f"Invalid git root: {repo_dir}")
+    await repo.checkout(commit_id)

metagpt/utils/git_repository.py

@@ -9,6 +9,8 @@
 from __future__ import annotations
 
 import shutil
+import subprocess
+import uuid
 from enum import Enum
 from pathlib import Path
 from typing import Dict, List
@@ -283,3 +285,36 @@ class GitRepository:
                 continue
             files.append(filename)
         return files
+
+    @classmethod
+    async def clone_from(cls, url: str | Path, output_dir: str | Path = None) -> "GitRepository":
+        from metagpt.context import Context
+
+        to_path = Path(output_dir or Path(__file__).parent / f"../../workspace/downloads/{uuid.uuid4().hex}").resolve()
+        to_path.mkdir(parents=True, exist_ok=True)
+        ctx = Context()
+        env = ctx.new_environ()
+        proxy = ["-c", f"http.proxy={ctx.config.proxy}"] if ctx.config.proxy else []
+        command = ["git", "clone"] + proxy + [str(url)]
+        logger.info(" ".join(command))
+        process = subprocess.Popen(command, cwd=str(to_path), stdout=subprocess.PIPE, stderr=subprocess.PIPE, env=env)
+        try:
+            # Wait for the process to complete, with a timeout
+            stdout, stderr = process.communicate(timeout=600)
+            info = f"{stdout.decode('utf-8')}\n{stderr.decode('utf-8')}"
+            logger.info(info)
+            dir_name = Path(url).with_suffix("").name
+            to_path = to_path / dir_name
+            if not cls.is_git_dir(to_path):
+                raise ValueError(info)
+            logger.info(f"git clone to {to_path}")
+            return GitRepository(local_path=to_path, auto_init=False)
+        except subprocess.TimeoutExpired:
+            logger.info("The command did not complete within the given timeout.")
+            process.kill()  # Kill the process if it times out
+            stdout, stderr = process.communicate()
+            raise ValueError(f"{stdout.decode('utf-8')}\n{stderr.decode('utf-8')}")
+
+    async def checkout(self, commit_id: str):
+        self._repository.git.checkout(commit_id)
+        logger.info(f"git checkout {commit_id}")

tests/metagpt/tools/libs/test_git.py

@@ -0,0 +1,31 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+
+import pytest
+from pydantic import BaseModel
+
+from metagpt.tools.libs.git import git_checkout, git_clone
+from metagpt.utils.git_repository import GitRepository
+
+
+class SWEBenchItem(BaseModel):
+    base_commit: str
+    repo: str
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize(
+    ["url", "commit_id"], [("https://github.com/sqlfluff/sqlfluff.git", "d19de0ecd16d298f9e3bfb91da122734c40c01e5")]
+)
+async def test_git(url: str, commit_id: str):
+    repo_dir = await git_clone(url)
+    assert repo_dir
+
+    await git_checkout(repo_dir, commit_id)
+
+    repo = GitRepository(repo_dir, auto_init=False)
+    repo.delete_repository()
+
+
+if __name__ == "__main__":
+    pytest.main([__file__, "-s"])