Search Engine Client

Lightweight, async wrappers around multiple search engine APIs (Serper/Google, DuckDuckGo, Bing). This package provides a small, opinionated interface for running text and image searches and returning standardized results.

Features

• Async-first search engine clients implementing a common BaseSearchEngine interface.
• Built-in engines: Serper (Google), DuckDuckGo, Bing (concrete implementations in src/search_wrapper/engines).
• Concurrent text + image fetching for engines that support both.

Installation

Install for development (editable) and add required dependencies:

pip install -e .
pip install httpx

Quick start

Set the required API key for services that need one (example for Serper/Google):

$env:GOOGLE_SERPER_API_KEY = "your_api_key_here"

Example usage (async):

import asyncio
from search_wrapper import GoogleSearchFromSerper

async def main():
	async with GoogleSearchFromSerper(api_key="YOUR_KEY") as client:
		results = await client.search("python async http client")
		print(results)

asyncio.run(main())

Alternatively import specific engine directly:

from search_wrapper.engines.serper import GoogleSearchFromSerper

Development notes

• Source layout lives under src/ (package name: search_wrapper). Either install the package in editable mode (pip install -e .) or add src to PYTHONPATH/VS Code python.analysis.extraPaths so imports resolve in the editor.
• The Serper client uses httpx.AsyncClient — ensure httpx is installed in your active environment.
• Tests and examples live at the project root (test_search.py, main.py).

Contributing

Feel free to open issues or PRs. When adding new engines, implement BaseSearchEngine and expose the engine in src/search_wrapper/__init__.py.