langchain-local-dev-loop

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

LangChain Local Dev Loop

LangChain本地开发循环

Overview

概述

Configure a rapid local development workflow for LangChain applications with testing, debugging, and hot reload capabilities.

为LangChain应用配置具备测试、调试和热重载能力的快速本地开发工作流。

Prerequisites

前提条件

Completed
```
langchain-install-auth
```
setup
Python 3.9+ with virtual environment
pytest and related testing tools
IDE with Python support (VS Code recommended)

已完成
```
langchain-install-auth
```
设置
Python 3.9+ 及虚拟环境
pytest及相关测试工具
支持Python的IDE（推荐VS Code）

Instructions

操作步骤

Step 1: Set Up Project Structure

步骤1：搭建项目结构

my-langchain-app/
├── src/
│   ├── __init__.py
│   ├── chains/
│   │   └── __init__.py
│   ├── agents/
│   │   └── __init__.py
│   └── prompts/
│       └── __init__.py
├── tests/
│   ├── __init__.py
│   ├── conftest.py
│   └── test_chains.py
├── .env
├── .env.example
├── pyproject.toml
└── README.md

my-langchain-app/
├── src/
│   ├── __init__.py
│   ├── chains/
│   │   └── __init__.py
│   ├── agents/
│   │   └── __init__.py
│   └── prompts/
│       └── __init__.py
├── tests/
│   ├── __init__.py
│   ├── conftest.py
│   └── test_chains.py
├── .env
├── .env.example
├── pyproject.toml
└── README.md

Step 2: Configure Testing

步骤2：配置测试

python

undefined

python

undefined

tests/conftest.py

import pytest from unittest.mock import MagicMock from langchain_core.messages import AIMessage

@pytest.fixture def mock_llm(): """Mock LLM for unit tests without API calls.""" mock = MagicMock() mock.invoke.return_value = AIMessage(content="Mocked response") return mock

@pytest.fixture def sample_prompt(): """Sample prompt for testing.""" from langchain_core.prompts import ChatPromptTemplate return ChatPromptTemplate.from_template("Test: {input}")

undefined

import pytest from unittest.mock import MagicMock from langchain_core.messages import AIMessage

@pytest.fixture def mock_llm(): """用于单元测试的Mock LLM，无需调用API。""" mock = MagicMock() mock.invoke.return_value = AIMessage(content="Mocked response") return mock

@pytest.fixture def sample_prompt(): """用于测试的示例提示词。""" from langchain_core.prompts import ChatPromptTemplate return ChatPromptTemplate.from_template("Test: {input}")

undefined

Step 3: Create Test File

步骤3：创建测试文件

python

undefined

python

undefined

tests/test_chains.py

def test_chain_construction(mock_llm, sample_prompt): """Test that chain can be constructed.""" from langchain_core.output_parsers import StrOutputParser

chain = sample_prompt | mock_llm | StrOutputParser()
assert chain is not None

def test_chain_invoke(mock_llm, sample_prompt): """Test chain invocation with mock.""" from langchain_core.output_parsers import StrOutputParser

chain = sample_prompt | mock_llm | StrOutputParser()
result = chain.invoke({"input": "test"})
assert result == "Mocked response"

undefined

def test_chain_construction(mock_llm, sample_prompt): """测试能否成功构建链。""" from langchain_core.output_parsers import StrOutputParser

chain = sample_prompt | mock_llm | StrOutputParser()
assert chain is not None

def test_chain_invoke(mock_llm, sample_prompt): """测试使用Mock调用链。""" from langchain_core.output_parsers import StrOutputParser

chain = sample_prompt | mock_llm | StrOutputParser()
result = chain.invoke({"input": "test"})
assert result == "Mocked response"

undefined

Step 4: Set Up Development Tools

步骤4：配置开发工具

toml

undefined

toml

undefined

pyproject.toml

[project] name = "my-langchain-app" version = "0.1.0" requires-python = ">=3.9" dependencies = [ "langchain>=0.3.0", "langchain-openai>=0.2.0", "python-dotenv>=1.0.0", ]

[project.optional-dependencies] dev = [ "pytest>=8.0.0", "pytest-asyncio>=0.23.0", "pytest-cov>=4.0.0", "ruff>=0.1.0", "mypy>=1.0.0", ]

[tool.pytest.ini_options] asyncio_mode = "auto" testpaths = ["tests"]

[tool.ruff] line-length = 100

undefined

[project] name = "my-langchain-app" version = "0.1.0" requires-python = ">=3.9" dependencies = [ "langchain>=0.3.0", "langchain-openai>=0.2.0", "python-dotenv>=1.0.0", ]

[project.optional-dependencies] dev = [ "pytest>=8.0.0", "pytest-asyncio>=0.23.0", "pytest-cov>=4.0.0", "ruff>=0.1.0", "mypy>=1.0.0", ]

[tool.pytest.ini_options] asyncio_mode = "auto" testpaths = ["tests"]

[tool.ruff] line-length = 100

undefined

Output

输出结果

Organized project structure with separation of concerns
pytest configuration with fixtures for mocking LLMs
Development dependencies configured
Ready for rapid iteration

职责分离的规范化项目结构
带有LLM模拟夹具的pytest配置
已配置的开发依赖
可立即进行快速迭代

Error Handling

错误处理

Error	Cause	Solution
Import Error	Missing package	Install with `pip install -e ".[dev]"`
Fixture Not Found	conftest.py issue	Ensure conftest.py is in tests/ directory
Async Test Error	Missing marker	Add `@pytest.mark.asyncio` decorator
Env Var Missing	.env not loaded	Use `python-dotenv` and load_dotenv()

错误类型	原因	解决方案
导入错误	缺失依赖包	使用 `pip install -e ".[dev]"` 安装
夹具未找到	conftest.py存在问题	确保conftest.py位于tests/目录下
异步测试错误	缺失标记	添加 `@pytest.mark.asyncio` 装饰器
环境变量缺失	.env文件未加载	使用 `python-dotenv` 并调用load_dotenv()

Examples

示例

Running Tests

运行测试

bash

undefined

bash

undefined

Run all tests

运行所有测试

pytest

Run with coverage

带覆盖率报告运行

pytest --cov=src --cov-report=html

Run specific test

运行指定测试

pytest tests/test_chains.py::test_chain_invoke -v

Watch mode (requires pytest-watch)

监听模式（需要pytest-watch）

ptw

undefined

ptw

undefined

Integration Test Example

集成测试示例

python

undefined

python

undefined

tests/test_integration.py

import pytest from dotenv import load_dotenv

load_dotenv()

@pytest.mark.integration def test_real_llm_call(): """Integration test with real LLM (requires API key).""" from langchain_openai import ChatOpenAI

llm = ChatOpenAI(model="gpt-4o-mini", temperature=0)
response = llm.invoke("Say 'test passed'")
assert "test" in response.content.lower()

undefined

import pytest from dotenv import load_dotenv

load_dotenv()

@pytest.mark.integration def test_real_llm_call(): """使用真实LLM的集成测试（需要API密钥）。""" from langchain_openai import ChatOpenAI

llm = ChatOpenAI(model="gpt-4o-mini", temperature=0)
response = llm.invoke("Say 'test passed'")
assert "test" in response.content.lower()

undefined

Resources

参考资源

Next Steps

后续步骤

Proceed to

langchain-sdk-patterns

for production-ready code patterns.

继续学习

langchain-sdk-patterns

以获取生产就绪的代码模式。