[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-deedy5--ddgs":3,"tool-deedy5--ddgs":61},[4,18,26,36,44,53],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":17},4358,"openclaw","openclaw\u002Fopenclaw","OpenClaw 是一款专为个人打造的本地化 AI 助手，旨在让你在自己的设备上拥有完全可控的智能伙伴。它打破了传统 AI 助手局限于特定网页或应用的束缚，能够直接接入你日常使用的各类通讯渠道，包括微信、WhatsApp、Telegram、Discord、iMessage 等数十种平台。无论你在哪个聊天软件中发送消息，OpenClaw 都能即时响应，甚至支持在 macOS、iOS 和 Android 设备上进行语音交互，并提供实时的画布渲染功能供你操控。\n\n这款工具主要解决了用户对数据隐私、响应速度以及“始终在线”体验的需求。通过将 AI 部署在本地，用户无需依赖云端服务即可享受快速、私密的智能辅助，真正实现了“你的数据，你做主”。其独特的技术亮点在于强大的网关架构，将控制平面与核心助手分离，确保跨平台通信的流畅性与扩展性。\n\nOpenClaw 非常适合希望构建个性化工作流的技术爱好者、开发者，以及注重隐私保护且不愿被单一生态绑定的普通用户。只要具备基础的终端操作能力（支持 macOS、Linux 及 Windows WSL2），即可通过简单的命令行引导完成部署。如果你渴望拥有一个懂你",349277,3,"2026-04-06T06:32:30",[13,14,15,16],"Agent","开发框架","图像","数据工具","ready",{"id":19,"name":20,"github_repo":21,"description_zh":22,"stars":23,"difficulty_score":10,"last_commit_at":24,"category_tags":25,"status":17},3808,"stable-diffusion-webui","AUTOMATIC1111\u002Fstable-diffusion-webui","stable-diffusion-webui 是一个基于 Gradio 构建的网页版操作界面，旨在让用户能够轻松地在本地运行和使用强大的 Stable Diffusion 图像生成模型。它解决了原始模型依赖命令行、操作门槛高且功能分散的痛点，将复杂的 AI 绘图流程整合进一个直观易用的图形化平台。\n\n无论是希望快速上手的普通创作者、需要精细控制画面细节的设计师，还是想要深入探索模型潜力的开发者与研究人员，都能从中获益。其核心亮点在于极高的功能丰富度：不仅支持文生图、图生图、局部重绘（Inpainting）和外绘（Outpainting）等基础模式，还独创了注意力机制调整、提示词矩阵、负向提示词以及“高清修复”等高级功能。此外，它内置了 GFPGAN 和 CodeFormer 等人脸修复工具，支持多种神经网络放大算法，并允许用户通过插件系统无限扩展能力。即使是显存有限的设备，stable-diffusion-webui 也提供了相应的优化选项，让高质量的 AI 艺术创作变得触手可及。",162132,"2026-04-05T11:01:52",[14,15,13],{"id":27,"name":28,"github_repo":29,"description_zh":30,"stars":31,"difficulty_score":32,"last_commit_at":33,"category_tags":34,"status":17},1381,"everything-claude-code","affaan-m\u002Feverything-claude-code","everything-claude-code 是一套专为 AI 编程助手（如 Claude Code、Codex、Cursor 等）打造的高性能优化系统。它不仅仅是一组配置文件，而是一个经过长期实战打磨的完整框架，旨在解决 AI 代理在实际开发中面临的效率低下、记忆丢失、安全隐患及缺乏持续学习能力等核心痛点。\n\n通过引入技能模块化、直觉增强、记忆持久化机制以及内置的安全扫描功能，everything-claude-code 能显著提升 AI 在复杂任务中的表现，帮助开发者构建更稳定、更智能的生产级 AI 代理。其独特的“研究优先”开发理念和针对 Token 消耗的优化策略，使得模型响应更快、成本更低，同时有效防御潜在的攻击向量。\n\n这套工具特别适合软件开发者、AI 研究人员以及希望深度定制 AI 工作流的技术团队使用。无论您是在构建大型代码库，还是需要 AI 协助进行安全审计与自动化测试，everything-claude-code 都能提供强大的底层支持。作为一个曾荣获 Anthropic 黑客大奖的开源项目，它融合了多语言支持与丰富的实战钩子（hooks），让 AI 真正成长为懂上",141543,2,"2026-04-06T11:32:54",[14,13,35],"语言模型",{"id":37,"name":38,"github_repo":39,"description_zh":40,"stars":41,"difficulty_score":32,"last_commit_at":42,"category_tags":43,"status":17},2271,"ComfyUI","Comfy-Org\u002FComfyUI","ComfyUI 是一款功能强大且高度模块化的视觉 AI 引擎，专为设计和执行复杂的 Stable Diffusion 图像生成流程而打造。它摒弃了传统的代码编写模式，采用直观的节点式流程图界面，让用户通过连接不同的功能模块即可构建个性化的生成管线。\n\n这一设计巧妙解决了高级 AI 绘图工作流配置复杂、灵活性不足的痛点。用户无需具备编程背景，也能自由组合模型、调整参数并实时预览效果，轻松实现从基础文生图到多步骤高清修复等各类复杂任务。ComfyUI 拥有极佳的兼容性，不仅支持 Windows、macOS 和 Linux 全平台，还广泛适配 NVIDIA、AMD、Intel 及苹果 Silicon 等多种硬件架构，并率先支持 SDXL、Flux、SD3 等前沿模型。\n\n无论是希望深入探索算法潜力的研究人员和开发者，还是追求极致创作自由度的设计师与资深 AI 绘画爱好者，ComfyUI 都能提供强大的支持。其独特的模块化架构允许社区不断扩展新功能，使其成为当前最灵活、生态最丰富的开源扩散模型工具之一，帮助用户将创意高效转化为现实。",107888,"2026-04-06T11:32:50",[14,15,13],{"id":45,"name":46,"github_repo":47,"description_zh":48,"stars":49,"difficulty_score":32,"last_commit_at":50,"category_tags":51,"status":17},4721,"markitdown","microsoft\u002Fmarkitdown","MarkItDown 是一款由微软 AutoGen 团队打造的轻量级 Python 工具，专为将各类文件高效转换为 Markdown 格式而设计。它支持 PDF、Word、Excel、PPT、图片（含 OCR）、音频（含语音转录）、HTML 乃至 YouTube 链接等多种格式的解析，能够精准提取文档中的标题、列表、表格和链接等关键结构信息。\n\n在人工智能应用日益普及的今天，大语言模型（LLM）虽擅长处理文本，却难以直接读取复杂的二进制办公文档。MarkItDown 恰好解决了这一痛点，它将非结构化或半结构化的文件转化为模型“原生理解”且 Token 效率极高的 Markdown 格式，成为连接本地文件与 AI 分析 pipeline 的理想桥梁。此外，它还提供了 MCP（模型上下文协议）服务器，可无缝集成到 Claude Desktop 等 LLM 应用中。\n\n这款工具特别适合开发者、数据科学家及 AI 研究人员使用，尤其是那些需要构建文档检索增强生成（RAG）系统、进行批量文本分析或希望让 AI 助手直接“阅读”本地文件的用户。虽然生成的内容也具备一定可读性，但其核心优势在于为机器",93400,"2026-04-06T19:52:38",[52,14],"插件",{"id":54,"name":55,"github_repo":56,"description_zh":57,"stars":58,"difficulty_score":10,"last_commit_at":59,"category_tags":60,"status":17},4487,"LLMs-from-scratch","rasbt\u002FLLMs-from-scratch","LLMs-from-scratch 是一个基于 PyTorch 的开源教育项目，旨在引导用户从零开始一步步构建一个类似 ChatGPT 的大型语言模型（LLM）。它不仅是同名技术著作的官方代码库，更提供了一套完整的实践方案，涵盖模型开发、预训练及微调的全过程。\n\n该项目主要解决了大模型领域“黑盒化”的学习痛点。许多开发者虽能调用现成模型，却难以深入理解其内部架构与训练机制。通过亲手编写每一行核心代码，用户能够透彻掌握 Transformer 架构、注意力机制等关键原理，从而真正理解大模型是如何“思考”的。此外，项目还包含了加载大型预训练权重进行微调的代码，帮助用户将理论知识延伸至实际应用。\n\nLLMs-from-scratch 特别适合希望深入底层原理的 AI 开发者、研究人员以及计算机专业的学生。对于不满足于仅使用 API，而是渴望探究模型构建细节的技术人员而言，这是极佳的学习资源。其独特的技术亮点在于“循序渐进”的教学设计：将复杂的系统工程拆解为清晰的步骤，配合详细的图表与示例，让构建一个虽小但功能完备的大模型变得触手可及。无论你是想夯实理论基础，还是为未来研发更大规模的模型做准备",90106,"2026-04-06T11:19:32",[35,15,13,14],{"id":62,"github_repo":63,"name":64,"description_en":65,"description_zh":66,"ai_summary_zh":67,"readme_en":68,"readme_zh":69,"quickstart_zh":70,"use_case_zh":71,"hero_image_url":72,"owner_login":73,"owner_name":74,"owner_avatar_url":75,"owner_bio":74,"owner_company":74,"owner_location":74,"owner_email":74,"owner_twitter":74,"owner_website":74,"owner_url":76,"languages":77,"stars":93,"forks":94,"last_commit_at":95,"license":96,"difficulty_score":97,"env_os":98,"env_gpu":99,"env_ram":99,"env_deps":100,"category_tags":106,"github_topics":108,"view_count":32,"oss_zip_url":74,"oss_zip_packed_at":74,"status":17,"created_at":116,"updated_at":117,"faqs":118,"releases":148},4615,"deedy5\u002Fddgs","ddgs","A metasearch library that aggregates results from diverse web search services","ddgs 是一款强大的开源元搜索库，旨在帮助开发者轻松聚合来自 Bing、Google、DuckDuckGo、Yahoo 等多个主流搜索引擎的搜索结果。它解决了单一搜索引擎覆盖不全、数据源受限以及手动编写爬虫维护成本高的问题，让用户只需一次调用，即可获取文本、图片、视频、新闻及书籍等多维度的全球网络信息。\n\n这款工具特别适合 Python 开发者、数据研究人员以及需要构建智能搜索功能的 AI 应用工程师使用。无论是为大型语言模型（LLM）搭建实时联网检索能力，还是开发需要多源数据验证的分析系统，ddgs 都能提供稳定支持。其独特亮点在于架构灵活：不仅提供了简洁的 Python 类库供代码直接集成，还内置了基于 FastAPI 的 API 服务器和兼容 MCP（Model Context Protocol）标准的服务端。这意味着它可以无缝对接 Cursor、Claude Desktop 等现代 AI 编程助手，让大模型具备“自主上网”搜索的能力。此外，ddgs 原生支持代理配置与超时控制，并采用懒加载机制优化性能，确保在复杂网络环境下依然高效可靠。通过统一的接口屏蔽不同搜索引擎的差异，d","ddgs 是一款强大的开源元搜索库，旨在帮助开发者轻松聚合来自 Bing、Google、DuckDuckGo、Yahoo 等多个主流搜索引擎的搜索结果。它解决了单一搜索引擎覆盖不全、数据源受限以及手动编写爬虫维护成本高的问题，让用户只需一次调用，即可获取文本、图片、视频、新闻及书籍等多维度的全球网络信息。\n\n这款工具特别适合 Python 开发者、数据研究人员以及需要构建智能搜索功能的 AI 应用工程师使用。无论是为大型语言模型（LLM）搭建实时联网检索能力，还是开发需要多源数据验证的分析系统，ddgs 都能提供稳定支持。其独特亮点在于架构灵活：不仅提供了简洁的 Python 类库供代码直接集成，还内置了基于 FastAPI 的 API 服务器和兼容 MCP（Model Context Protocol）标准的服务端。这意味着它可以无缝对接 Cursor、Claude Desktop 等现代 AI 编程助手，让大模型具备“自主上网”搜索的能力。此外，ddgs 原生支持代理配置与超时控制，并采用懒加载机制优化性能，确保在复杂网络环境下依然高效可靠。通过统一的接口屏蔽不同搜索引擎的差异，ddgs 让获取全球信息变得简单而规范。","![Python >= 3.10](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fpython->=3.10-red.svg) [![](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdeedy5_ddgs_readme_d452f86bb2ad.png)](https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Freleases) [![](https:\u002F\u002Fbadge.fury.io\u002Fpy\u002Fddgs.svg)](https:\u002F\u002Fpypi.org\u002Fproject\u002Fddgs)\n# DDGS | Dux Distributed Global Search\u003Ca name=\"TOP\">\u003C\u002Fa>\n\nA metasearch library that aggregates results from diverse web search services.\n\n\n## Table of Contents\n* [Install](#install)\n* [CLI version](#cli-version)\n* [API Server](#api-server)\n* [MCP Server](#mcp-server)\n* [Engines](#engines)\n* [DDGS class](#ddgs-class)\n* [1. text()](#1-text)\n* [2. images()](#2-images)\n* [3. videos()](#3-videos)\n* [4. news()](#4-news)\n* [5. books()](#5-books)\n* [6. extract()](#6-extract)\n* [Disclaimer](#disclaimer)\n\n___\n## Install\n```python\npip install -U ddgs       # Base install\npip install -U ddgs[api]  # API server (FastAPI)\npip install -U ddgs[mcp]  # MCP server (stdio)\n```\n\n## CLI version\n\n```python3\nddgs --help\n```\n\n[Go To TOP](#TOP)\n___\n\n## API Server\n\n- **Install**\n```bash\npip install -U ddgs[api]\n```\n\n- **CLI**\n```bash\nddgs api              # Start server in foreground\nddgs api -d           # Start in detached mode (background)\nddgs api -s           # Stop detached server\nddgs api --host 127.0.0.1 --port 9000  # Custom host\u002Fport\nddgs api -pr socks5h:\u002F\u002F127.0.0.1:9150  # With proxy\n```\n\n- **Docker compose**\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs && cd ddgs\ndocker-compose up --build\n```\n\n- **Bash script**\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs && cd ddgs\nchmod +x start_api.sh\n.\u002Fstart_api.sh\n```\n\n#### Endpoints\n\n| Endpoint | Method | Description |\n|----------|--------|-------------|\n| `\u002Fsearch\u002Ftext` | GET, POST | Text search |\n| `\u002Fsearch\u002Fimages` | GET, POST | Image search |\n| `\u002Fsearch\u002Fnews` | GET, POST | News search |\n| `\u002Fsearch\u002Fvideos` | GET, POST | Video search |\n| `\u002Fsearch\u002Fbooks` | GET, POST | Book search |\n| `\u002Fextract` | GET, POST | Extract content from URL |\n| `\u002Fhealth` | GET | Health check |\n| `\u002Fdocs` | GET | Swagger UI |\n| `\u002Fredoc` | GET | ReDoc documentation |\n\n[Go To TOP](#TOP)\n___\n\n## MCP Server\n\n- **Install**\n```bash\npip install -U ddgs[mcp]\n```\n\n- **CLI**\n```bash\nddgs mcp    # Start MCP server (stdio transport)\nddgs mcp -pr socks5h:\u002F\u002F127.0.0.1:9150  # With proxy\n```\n\n#### Available Tools\n\n| Tool | Description |\n|------|-------------|\n| `search_text` | Web text search |\n| `search_images` | Image search |\n| `search_news` | News search |\n| `search_videos` | Video search |\n| `search_books` | Book search |\n| `extract_content` | Extract content from a URL |\n\n#### Client Configuration\n\nFor MCP clients like Cursor or Claude Desktop:\n```json\n{\n  \"mcpServers\": {\n    \"ddgs\": {\n      \"command\": \"ddgs\",\n      \"args\": [\"mcp\"]\n    }\n  }\n}\n```\n\n[Go To TOP](#TOP)\n___\n\n## Engines\n\n| DDGS function | Available backends |\n| --------------|:-------------------|\n| text()        | `bing`, `brave`, `duckduckgo`, `google`, `grokipedia`, `mojeek`, `yandex`, `yahoo`, `wikipedia`|\n| images()      | `bing`, `duckduckgo` |\n| videos()      | `duckduckgo` |\n| news()        | `bing`, `duckduckgo`, `yahoo` |\n| books()       | `annasarchive` |\n\n[Go To TOP](#TOP)\n\n## DDGS class\n\nDDGS class is lazy-loaded.\n\n```python3\nclass DDGS:\n    \"\"\"Dux Distributed Global Search. A metasearch library that aggregates results from diverse web search services.\n\n    Args:\n        proxy (str, optional): proxy for the HTTP client, supports http\u002Fhttps\u002Fsocks5 protocols.\n            example: \"http:\u002F\u002Fuser:pass@example.com:3128\". Defaults to None.\n        timeout (int, optional): Timeout value for the HTTP client. Defaults to 5.\n        verify: (bool | str):  True to verify, False to skip, or a str path to a PEM file. Defaults to True.\n    \"\"\"\n```\n\nHere is an example of initializing the DDGS class.\n```python3\nfrom ddgs import DDGS\n\nresults = DDGS().text(\"python programming\", max_results=5)\nprint(results)\n```\n\n[Go To TOP](#TOP)\n\n## 1. text()\n\n```python\ndef text(\n    query: str,\n    region: str = \"us-en\",\n    safesearch: str = \"moderate\",\n    timelimit: str | None = None,\n    max_results: int | None = 10,\n    page: int = 1,\n    backend: str = \"auto\",\n) -> list[dict[str, str]]:\n    \"\"\"DDGS text metasearch.\n\n    Args:\n        query: text search query.\n        region: us-en, uk-en, ru-ru, etc. Defaults to us-en.\n        safesearch: on, moderate, off. Defaults to \"moderate\".\n        timelimit: d, w, m, y. Defaults to None.\n        max_results: maximum number of results. Defaults to 10.\n        page: page of results. Defaults to 1.\n        backend: A single or comma-delimited backends. Defaults to \"auto\".\n\n    Returns:\n        List of dictionaries with search results.\n    \"\"\"\n```\n***Example***\n```python\nresults = DDGS().text('live free or die', region='us-en', safesearch='off', timelimit='y', page=1, backend=\"auto\")\n# Searching for pdf files\nresults = DDGS().text('russia filetype:pdf', region='us-en', safesearch='off', timelimit='y', page=1, backend=\"auto\")\nprint(results)\n[\n    {\n        \"title\": \"News, sport, celebrities and gossip | The Sun\",\n        \"href\": \"https:\u002F\u002Fwww.thesun.co.uk\u002F\",\n        \"body\": \"Get the latest news, exclusives, sport, celebrities, showbiz, politics, business and lifestyle from The Sun\",\n    }, ...\n]\n```\n\n[Go To TOP](#TOP)\n\n## 2. images()\n\n```python\ndef images(\n    query: str,\n    region: str = \"us-en\",\n    safesearch: str = \"moderate\",\n    timelimit: str | None = None,\n    max_results: int | None = 10,\n    page: int = 1,\n    backend: str = \"auto\",\n    size: str | None = None,\n    color: str | None = None,\n    type_image: str | None = None,\n    layout: str | None = None,\n    license_image: str | None = None,\n) -> list[dict[str, str]]:\n    \"\"\"DDGS images metasearch.\n\n    Args:\n        query: images search query.\n        region: us-en, uk-en, ru-ru, etc. Defaults to us-en.\n        safesearch: on, moderate, off. Defaults to \"moderate\".\n        timelimit: d, w, m, y. Defaults to None.\n        max_results: maximum number of results. Defaults to 10.\n        page: page of results. Defaults to 1.\n        backend: A single or comma-delimited backends. Defaults to \"auto\".\n        size: Small, Medium, Large, Wallpaper. Defaults to None.\n        color: color, Monochrome, Red, Orange, Yellow, Green, Blue,\n            Purple, Pink, Brown, Black, Gray, Teal, White. Defaults to None.\n        type_image: photo, clipart, gif, transparent, line.\n            Defaults to None.\n        layout: Square, Tall, Wide. Defaults to None.\n        license_image: any (All Creative Commons), Public (PublicDomain),\n            Share (Free to Share and Use), ShareCommercially (Free to Share and Use Commercially),\n            Modify (Free to Modify, Share, and Use), ModifyCommercially (Free to Modify, Share, and\n            Use Commercially). Defaults to None.\n\n    Returns:\n        List of dictionaries with images search results.\n    \"\"\"\n```\n***Example***\n```python\nresults = DDGS().images(\n    query=\"butterfly\",\n    region=\"us-en\",\n    safesearch=\"off\",\n    timelimit=\"m\",\n    page=1,\n    backend=\"auto\",\n    size=None,\n    color=\"Monochrome\",\n    type_image=None,\n    layout=None,\n    license_image=None,\n)\nprint(images)\n[\n    {\n        \"title\": \"File:The Sun by the Atmospheric Imaging Assembly of NASA's Solar ...\",\n        \"image\": \"https:\u002F\u002Fupload.wikimedia.org\u002Fwikipedia\u002Fcommons\u002Fb\u002Fb4\u002FThe_Sun_by_the_Atmospheric_Imaging_Assembly_of_NASA's_Solar_Dynamics_Observatory_-_20100819.jpg\",\n        \"thumbnail\": \"https:\u002F\u002Ftse4.mm.bing.net\u002Fth?id=OIP.lNgpqGl16U0ft3rS8TdFcgEsEe&pid=Api\",\n        \"url\": \"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FFile:The_Sun_by_the_Atmospheric_Imaging_Assembly_of_NASA's_Solar_Dynamics_Observatory_-_20100819.jpg\",\n        \"height\": 3860,\n        \"width\": 4044,\n        \"source\": \"Bing\",\n    }, ...\n]\n```\n\n[Go To TOP](#TOP)\n\n## 3. videos()\n\n```python\ndef videos(\n    query: str,\n    region: str = \"us-en\",\n    safesearch: str = \"moderate\",\n    timelimit: str | None = None,\n    max_results: int | None = 10,\n    page: int = 1,\n    backend: str = \"auto\",\n    resolution: str | None = None,\n    duration: str | None = None,\n    license_videos: str | None = None,\n) -> list[dict[str, str]]:\n    \"\"\"DDGS videos metasearch.\n\n    Args:\n        query: videos search query.\n        region: us-en, uk-en, ru-ru, etc. Defaults to us-en.\n        safesearch: on, moderate, off. Defaults to \"moderate\".\n        timelimit: d, w, m. Defaults to None.\n        max_results: maximum number of results. Defaults to 10.\n        page: page of results. Defaults to 1.\n        backend: A single or comma-delimited backends. Defaults to \"auto\".\n        resolution: high, standart. Defaults to None.\n        duration: short, medium, long. Defaults to None.\n        license_videos: creativeCommon, youtube. Defaults to None.\n\n    Returns:\n        List of dictionaries with videos search results.\n    \"\"\"\n```\n***Example***\n```python\nresults = DDGS().videos(\n    query=\"cars\",\n    region=\"us-en\",\n    safesearch=\"off\",\n    timelimit=\"w\",\n    page=1,\n    backend=\"auto\",\n    resolution=\"high\",\n    duration=\"medium\",\n)\nprint(results)\n[\n    {\n        \"content\": \"https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=6901-C73P3g\",\n        \"description\": \"Watch the Best Scenes of popular Tamil Serial #Meena that airs on Sun TV. Watch all Sun TV serials immediately after the TV telecast on Sun NXT app. *Free for Indian Users only Download here: Android - http:\u002F\u002Fbit.ly\u002FSunNxtAdroid iOS: India - http:\u002F\u002Fbit.ly\u002FsunNXT Watch on the web - https:\u002F\u002Fwww.sunnxt.com\u002F Two close friends, Chidambaram ...\",\n        \"duration\": \"8:22\",\n        \"embed_html\": '\u003Ciframe width=\"1280\" height=\"720\" src=\"https:\u002F\u002Fwww.youtube.com\u002Fembed\u002F6901-C73P3g?autoplay=1\" frameborder=\"0\" allowfullscreen>\u003C\u002Fiframe>',\n        \"embed_url\": \"https:\u002F\u002Fwww.youtube.com\u002Fembed\u002F6901-C73P3g?autoplay=1\",\n        \"image_token\": \"6c070b5f0e24e5972e360d02ddeb69856202f97718ea6c5d5710e4e472310fa3\",\n        \"images\": {\n            \"large\": \"https:\u002F\u002Ftse4.mm.bing.net\u002Fth?id=OVF.JWBFKm1u%2fHd%2bz2e1GitsQw&pid=Api\",\n            \"medium\": \"https:\u002F\u002Ftse4.mm.bing.net\u002Fth?id=OVF.JWBFKm1u%2fHd%2bz2e1GitsQw&pid=Api\",\n            \"motion\": \"\",\n            \"small\": \"https:\u002F\u002Ftse4.mm.bing.net\u002Fth?id=OVF.JWBFKm1u%2fHd%2bz2e1GitsQw&pid=Api\",\n        },\n        \"provider\": \"Bing\",\n        \"published\": \"2024-07-03T05:30:03.0000000\",\n        \"publisher\": \"YouTube\",\n        \"statistics\": {\"viewCount\": 29059},\n        \"title\": \"Meena - Best Scenes | 02 July 2024 | Tamil Serial | Sun TV\",\n        \"uploader\": \"Sun TV\",\n    }, ...\n]\n```\n\n[Go To TOP](#TOP)\n\n## 4. news()\n\n```python\ndef news(\n    query: str,\n    region: str = \"us-en\",\n    safesearch: str = \"moderate\",\n    timelimit: str | None = None,\n    max_results: int | None = 10,\n    page: int = 1,\n    backend: str = \"auto\",\n) -> list[dict[str, str]]:\n    \"\"\"DDGS news metasearch.\n\n    Args:\n        query: news search query.\n        region: us-en, uk-en, ru-ru, etc. Defaults to us-en.\n        safesearch: on, moderate, off. Defaults to \"moderate\".\n        timelimit: d, w, m. Defaults to None.\n        max_results: maximum number of results. Defaults to 10.\n        page: page of results. Defaults to 1.\n        backend: A single or comma-delimited backends. Defaults to \"auto\".\n\n    Returns:\n        List of dictionaries with news search results.\n    \"\"\"\n```\n***Example***\n```python\nresults = DDGS().news(query=\"sun\", region=\"us-en\", safesearch=\"off\", timelimit=\"m\", page=1, backend=\"auto\")\nprint(results)\n[\n    {\n        \"date\": \"2024-07-03T16:25:22+00:00\",\n        \"title\": \"Murdoch's Sun Endorses Starmer's Labour Day Before UK Vote\",\n        \"body\": \"Rupert Murdoch's Sun newspaper endorsed Keir Starmer and his opposition Labour Party to win the UK general election, a dramatic move in the British media landscape that illustrates the country's shifting political sands.\",\n        \"url\": \"https:\u002F\u002Fwww.msn.com\u002Fen-us\u002Fmoney\u002Fother\u002Fmurdoch-s-sun-endorses-starmer-s-labour-day-before-uk-vote\u002Far-BB1plQwl\",\n        \"image\": \"https:\u002F\u002Fimg-s-msn-com.akamaized.net\u002Ftenant\u002Famp\u002Fentityid\u002FBB1plZil.img?w=2000&h=1333&m=4&q=79\",\n        \"source\": \"Bloomberg on MSN.com\",\n    }, ...\n]\n```\n\n[Go To TOP](#TOP)\n\n## 5. books()\n\n```python\ndef books(\n    query: str,\n    max_results: int | None = 10,\n    page: int = 1,\n    backend: str = \"auto\",\n) -> list[dict[str, str]]:\n    \"\"\"DDGS books metasearch.\n\n    Args:\n        query: news search query.\n        max_results: maximum number of results. Defaults to 10.\n        page: page of results. Defaults to 1.\n        backend: A single or comma-delimited backends. Defaults to \"auto\".\n\n    Returns:\n        List of dictionaries with news search results.\n    \"\"\"\n```\n***Example***\n```python\nresults = DDGS().books(query=\"sea wolf jack london\", page=1, backend=\"auto\")\nprint(results)\n[\n    {\n        'title': 'The Sea-Wolf',\n        'author': 'Jack London',\n        'publisher': 'DigiCat, 2022',\n        'info': 'English [en], .epub, 🚀\u002Fzlib, 0.5MB, 📗 Book (unknown)',\n        'url': 'https:\u002F\u002Fannas-archive.li\u002Fmd5\u002F574f6556f1df6717de4044e36c7c2782',\n        'thumbnail': 'https:\u002F\u002Fs3proxy.cdn-zlib.sk\u002F\u002Fcovers299\u002Fcollections\u002Fuserbooks\u002Fda4954486be7c2b2b9f70b2aa5bcf01292de3ea510b5656f892821950ded9ada.jpg',\n    }, ...\n]\n```\n\n[Go To TOP](#TOP)\n\n## 6. extract()\n\nFetch a URL and extract its content in various formats.\n\n```python\ndef extract(\n    url: str,\n    fmt: str = \"text_markdown\",\n) -> dict[str, str | bytes]:\n    \"\"\"Fetch a URL and extract its content.\n\n    Args:\n        url: The URL to fetch and extract content from.\n        fmt: Output format:\n            \"text_markdown\" (HTML→Markdown, preserves links\u002Fheaders\u002Flists),\n            \"text_plain\" (HTML→plain text),\n            \"text_rich\" (HTML→rich text with headers\u002Flists),\n            \"text\" (raw HTML),\n            \"content\" (raw bytes).\n\n    Returns:\n        Dictionary with 'url' and 'content' keys.\n    \"\"\"\n```\n***Examples***\n```python\n# Markdown (default) - preserves links, headers, lists\nresult = DDGS().extract(\"https:\u002F\u002Fexample.com\")\nprint(result)\n{\"url\": \"https:\u002F\u002Fexample.com\", \"content\": \"# Example Domain\\n\\nThis domain is for use in...\"}\n\n# Plain text\nresult = DDGS().extract(\"https:\u002F\u002Fexample.com\", fmt=\"text_plain\")\n\n# Rich text (headers\u002Flists, no link URLs)\nresult = DDGS().extract(\"https:\u002F\u002Fexample.com\", fmt=\"text_rich\")\n\n# Raw HTML\nresult = DDGS().extract(\"https:\u002F\u002Fexample.com\", fmt=\"text\")\n\n# Raw bytes\nresult = DDGS().extract(\"https:\u002F\u002Fexample.com\", fmt=\"content\")\n```\n\n***CLI***\n```bash\nddgs extract -u https:\u002F\u002Fexample.com\nddgs extract -u https:\u002F\u002Fexample.com -f text_plain\nddgs extract -u https:\u002F\u002Fexample.com -f content -o output.json\n```\n\n[Go To TOP](#TOP)\n\n## Disclaimer\n\nThis library is for educational purposes only.\n","![Python >= 3.10](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fpython->=3.10-red.svg) [![](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdeedy5_ddgs_readme_d452f86bb2ad.png)](https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Freleases) [![](https:\u002F\u002Fbadge.fury.io\u002Fpy\u002Fddgs.svg)](https:\u002F\u002Fpypi.org\u002Fproject\u002Fddgs)\n# DDGS | 杜克斯分布式全球搜索\u003Ca name=\"TOP\">\u003C\u002Fa>\n\n一个元搜索引擎库，可聚合来自不同网络搜索服务的结果。\n\n\n## 目录\n* [安装](#install)\n* [CLI 版本](#cli-version)\n* [API 服务器](#api-server)\n* [MCP 服务器](#mcp-server)\n* [引擎](#engines)\n* [DDGS 类](#ddgs-class)\n* [1. text()](#1-text)\n* [2. images()](#2-images)\n* [3. videos()](#3-videos)\n* [4. news()](#4-news)\n* [5. books()](#5-books)\n* [6. extract()](#6-extract)\n* [免责声明](#disclaimer)\n\n___\n## 安装\n```python\npip install -U ddgs       # 基础安装\npip install -U ddgs[api]  # API 服务器 (FastAPI)\npip install -U ddgs[mcp]  # MCP 服务器 (stdio)\n```\n\n## CLI 版本\n\n```python3\nddgs --help\n```\n\n[返回顶部](#TOP)\n___\n\n## API 服务器\n\n- **安装**\n```bash\npip install -U ddgs[api]\n```\n\n- **CLI**\n```bash\nddgs api              # 在前台启动服务器\nddgs api -d           # 以分离模式（后台）启动\nddgs api -s           # 停止分离模式的服务器\nddgs api --host 127.0.0.1 --port 9000  # 自定义主机\u002F端口\nddgs api -pr socks5h:\u002F\u002F127.0.0.1:9150  # 使用代理\n```\n\n- **Docker Compose**\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs && cd ddgs\ndocker-compose up --build\n```\n\n- **Bash 脚本**\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs && cd ddgs\nchmod +x start_api.sh\n.\u002Fstart_api.sh\n```\n\n#### 端点\n\n| 端点 | 方法 | 描述 |\n|----------|--------|-------------|\n| `\u002Fsearch\u002Ftext` | GET, POST | 文本搜索 |\n| `\u002Fsearch\u002Fimages` | GET, POST | 图片搜索 |\n| `\u002Fsearch\u002Fnews` | GET, POST | 新闻搜索 |\n| `\u002Fsearch\u002Fvideos` | GET, POST | 视频搜索 |\n| `\u002Fsearch\u002Fbooks` | GET, POST | 书籍搜索 |\n| `\u002Fextract` | GET, POST | 从 URL 提取内容 |\n| `\u002Fhealth` | GET | 健康检查 |\n| `\u002Fdocs` | GET | Swagger UI |\n| `\u002Fredoc` | GET | ReDoc 文档 |\n\n[返回顶部](#TOP)\n___\n\n## MCP 服务器\n\n- **安装**\n```bash\npip install -U ddgs[mcp]\n```\n\n- **CLI**\n```bash\nddgs mcp    # 启动 MCP 服务器（stdio 传输）\nddgs mcp -pr socks5h:\u002F\u002F127.0.0.1:9150  # 使用代理\n```\n\n#### 可用工具\n\n| 工具 | 描述 |\n|------|-------------|\n| `search_text` | 网页文本搜索 |\n| `search_images` | 图片搜索 |\n| `search_news` | 新闻搜索 |\n| `search_videos` | 视频搜索 |\n| `search_books` | 书籍搜索 |\n| `extract_content` | 从 URL 提取内容 |\n\n#### 客户端配置\n\n对于 Cursor 或 Claude Desktop 等 MCP 客户端：\n```json\n{\n  \"mcpServers\": {\n    \"ddgs\": {\n      \"command\": \"ddgs\",\n      \"args\": [\"mcp\"]\n    }\n  }\n}\n```\n\n[返回顶部](#TOP)\n___\n\n## 引擎\n\n| DDGS 函数 | 可用后端 |\n| --------------|:-------------------|\n| text()        | `bing`, `brave`, `duckduckgo`, `google`, `grokipedia`, `mojeek`, `yandex`, `yahoo`, `wikipedia`|\n| images()      | `bing`, `duckduckgo` |\n| videos()      | `duckduckgo` |\n| news()        | `bing`, `duckduckgo`, `yahoo` |\n| books()       | `annasarchive` |\n\n[返回顶部](#TOP)\n\n## DDGS 类\n\nDDGS 类采用懒加载方式。\n\n```python3\nclass DDGS:\n    \"\"\"杜克斯分布式全球搜索。一个元搜索引擎库，可聚合来自不同网络搜索服务的结果。\n\n    参数:\n        proxy (str, 可选): HTTP 客户端使用的代理，支持 http\u002Fhttps\u002Fsocks5 协议。\n            示例: \"http:\u002F\u002Fuser:pass@example.com:3128\"。默认为 None。\n        timeout (int, 可选): HTTP 客户端的超时时间。默认为 5 秒。\n        verify: (bool 或 str): True 表示验证，False 表示跳过，或指定 PEM 文件路径。默认为 True。\n    \"\"\"\n```\n\n以下是初始化 DDGS 类的示例。\n```python3\nfrom ddgs import DDGS\n\nresults = DDGS().text(\"python 编程\", max_results=5)\nprint(results)\n```\n\n[返回顶部](#TOP)\n\n## 1. text()\n\n```python\ndef text(\n    query: str,\n    region: str = \"us-en\",\n    safesearch: str = \"moderate\",\n    timelimit: str | None = None,\n    max_results: int | None = 10,\n    page: int = 1,\n    backend: str = \"auto\",\n) -> list[dict[str, str]]:\n    \"\"\"DDGS 文本元搜索引擎。\n\n    参数:\n        query: 文本搜索查询。\n        region: us-en、uk-en、ru-ru 等。默认为 us-en。\n        safesearch: on、moderate、off。默认为 \"moderate\"。\n        timelimit: d、w、m、y。默认为 None。\n        max_results: 最大结果数。默认为 10。\n        page: 结果页码。默认为 1。\n        backend: 单个或逗号分隔的后端列表。默认为 \"auto\"。\n\n    返回:\n        包含搜索结果的字典列表。\n    \"\"\"\n```\n***示例***\n```python\nresults = DDGS().text('live free or die', region='us-en', safesearch='off', timelimit='y', page=1, backend=\"auto\")\n# 搜索 PDF 文件\nresults = DDGS().text('russia filetype:pdf', region='us-en', safesearch='off', timelimit='y', page=1, backend=\"auto\")\nprint(results)\n[\n    {\n        \"title\": \"新闻、体育、名人与八卦 | 太阳报\",\n        \"href\": \"https:\u002F\u002Fwww.thesun.co.uk\u002F\",\n        \"body\": \"获取太阳报最新的新闻、独家报道、体育、名人、娱乐、政治、商业和生活方式资讯\",\n    }, ...\n]\n```\n\n[返回顶部](#TOP)\n\n## 2. images()\n\n```python\ndef images(\n    query: str,\n    region: str = \"us-en\",\n    safesearch: str = \"moderate\",\n    timelimit: str | None = None,\n    max_results: int | None = 10,\n    page: int = 1,\n    backend: str = \"auto\",\n    size: str | None = None,\n    color: str | None = None,\n    type_image: str | None = None,\n    layout: str | None = None,\n    license_image: str | None = None,\n) -> list[dict[str, str]]:\n    \"\"\"DDGS 图片元搜索引擎。\n\n    参数:\n        query: 图片搜索关键词。\n        region: 美国-英语、英国-英语、俄罗斯-俄语等。默认为美国-英语。\n        safesearch: 开启、中等、关闭。默认为“中等”。\n        timelimit: d（天）、w（周）、m（月）、y（年）。默认为 None。\n        max_results: 最大结果数。默认为 10。\n        page: 结果页码。默认为 1。\n        backend: 单个或以逗号分隔的搜索引擎。默认为“auto”。\n        size: 小、中、大、壁纸。默认为 None。\n        color: 彩色、单色、红色、橙色、黄色、绿色、蓝色、紫色、粉色、棕色、黑色、灰色、青绿色、白色。默认为 None。\n        type_image: 照片、剪贴画、GIF、透明、线稿。默认为 None。\n        layout: 正方形、长图、宽图。默认为 None。\n        license_image: 任意（所有知识共享）、公共领域、可分享、可商业使用、可修改、可商业修改。默认为 None。\n\n    返回:\n        包含图片搜索结果的字典列表。\n    \"\"\"\n```\n***示例***\n```python\nresults = DDGS().images(\n    query=\"butterfly\",\n    region=\"us-en\",\n    safesearch=\"off\",\n    timelimit=\"m\",\n    page=1,\n    backend=\"auto\",\n    size=None,\n    color=\"Monochrome\",\n    type_image=None,\n    layout=None,\n    license_image=None,\n)\nprint(images)\n[\n    {\n        \"title\": \"文件：太阳由美国宇航局太阳动力学天文台的大气成像组件拍摄……\",\n        \"image\": \"https:\u002F\u002Fupload.wikimedia.org\u002Fwikipedia\u002Fcommons\u002Fb\u002Fb4\u002FThe_Sun_by_the_Atmospheric_Imaging_Assembly_of_NASA's_Solar_Dynamics_Observatory_-_20100819.jpg\",\n        \"thumbnail\": \"https:\u002F\u002Ftse4.mm.bing.net\u002Fth?id=OIP.lNgpqGl16U0ft3rS8TdFcgEsEe&pid=Api\",\n        \"url\": \"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FFile:The_Sun_by_the_Atmospheric_Imaging_Assembly_of_NASA's_Solar_Dynamics_Observatory_-_20100819.jpg\",\n        \"height\": 3860,\n        \"width\": 4044,\n        \"source\": \"Bing\",\n    }, ...\n]\n```\n\n[返回顶部](#TOP)\n\n## 3. videos()\n\n```python\ndef videos(\n    query: str,\n    region: str = \"us-en\",\n    safesearch: str = \"moderate\",\n    timelimit: str | None = None,\n    max_results: int | None = 10,\n    page: int = 1,\n    backend: str = \"auto\",\n    resolution: str | None = None,\n    duration: str | None = None,\n    license_videos: str | None = None,\n) -> list[dict[str, str]]:\n    \"\"\"DDGS 视频元搜索引擎。\n\n    参数:\n        query: 视频搜索关键词。\n        region: 美国-英语、英国-英语、俄罗斯-俄语等。默认为美国-英语。\n        safesearch: 开启、中等、关闭。默认为“中等”。\n        timelimit: d（天）、w（周）、m（月）。默认为 None。\n        max_results: 最大结果数。默认为 10。\n        page: 结果页码。默认为 1。\n        backend: 单个或以逗号分隔的搜索引擎。默认为“auto”。\n        resolution: 高清、标准。默认为 None。\n        duration: 短、中、长。默认为 None。\n        license_videos: 知识共享、YouTube。默认为 None。\n\n    返回:\n        包含视频搜索结果的字典列表。\n    \"\"\"\n```\n***示例***\n```python\nresults = DDGS().videos(\n    query=\"cars\",\n    region=\"us-en\",\n    safesearch=\"off\",\n    timelimit=\"w\",\n    page=1,\n    backend=\"auto\",\n    resolution=\"high\",\n    duration=\"medium\",\n)\nprint(results)\n[\n    {\n        \"content\": \"https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=6901-C73P3g\",\n        \"description\": \"观看热门泰米尔连续剧《Meena》在Sun TV播出的最佳片段。在Sun NXT应用上，您可以在电视播出后立即观看所有Sun TV连续剧。*仅限印度用户免费下载：Android - http:\u002F\u002Fbit.ly\u002FSunNxtAdroid iOS：印度 - http:\u002F\u002Fbit.ly\u002FsunNXT 在网页上观看 - https:\u002F\u002Fwww.sunnxt.com\u002F 两位亲密的朋友，Chidambaram ...\",\n        \"duration\": \"8:22\",\n        \"embed_html\": '\u003Ciframe width=\"1280\" height=\"720\" src=\"https:\u002F\u002Fwww.youtube.com\u002Fembed\u002F6901-C73P3g?autoplay=1\" frameborder=\"0\" allowfullscreen>\u003C\u002Fiframe>',\n        \"embed_url\": \"https:\u002F\u002Fwww.youtube.com\u002Fembed\u002F6901-C73P3g?autoplay=1\",\n        \"image_token\": \"6c070b5f0e24e5972e360d02ddeb69856202f97718ea6c5d5710e4e472310fa3\",\n        \"images\": {\n            \"large\": \"https:\u002F\u002Ftse4.mm.bing.net\u002Fth?id=OVF.JWBFKm1u%2fHd%2bz2e1GitsQw&pid=Api\",\n            \"medium\": \"https:\u002F\u002Ftse4.mm.bing.net\u002Fth?id=OVF.JWBFKm1u%2fHd%2bz2e1GitsQw&pid=Api\",\n            \"motion\": \"\",\n            \"small\": \"https:\u002F\u002Ftse4.mm.bing.net\u002Fth?id=OVF.JWBFKm1u%2fHd%2bz2e1GitsQw&pid=Api\",\n        },\n        \"provider\": \"Bing\",\n        \"published\": \"2024-07-03T05:30:03.0000000\",\n        \"publisher\": \"YouTube\",\n        \"statistics\": {\"viewCount\": 29059},\n        \"title\": \"Meena - 最佳片段 | 2024年7月2日 | 泰米尔连续剧 | Sun TV\",\n        \"uploader\": \"Sun TV\",\n    }, ...\n]\n```\n\n[返回顶部](#TOP)\n\n## 4. news()\n\n```python\ndef news(\n    query: str,\n    region: str = \"us-en\",\n    safesearch: str = \"moderate\",\n    timelimit: str | None = None,\n    max_results: int | None = 10,\n    page: int = 1,\n    backend: str = \"auto\",\n) -> list[dict[str, str]]:\n    \"\"\"DDGS 新闻元搜索引擎。\n\n    参数:\n        query: 新闻搜索关键词。\n        region: 美国-英语、英国-英语、俄罗斯-俄语等。默认为美国-英语。\n        safesearch: 开启、中等、关闭。默认为“中等”。\n        timelimit: d（天）、w（周）、m（月）。默认为 None。\n        max_results: 最大结果数。默认为 10。\n        page: 结果页码。默认为 1。\n        backend: 单个或以逗号分隔的搜索引擎。默认为“auto”。\n\n    返回:\n        包含新闻搜索结果的字典列表。\n    \"\"\"\n```\n***示例***\n```python\nresults = DDGS().news(query=\"sun\", region=\"us-en\", safesearch=\"off\", timelimit=\"m\", page=1, backend=\"auto\")\nprint(results)\n[\n    {\n        \"date\": \"2024-07-03T16:25:22+00:00\",\n        \"title\": \"默多克旗下的《太阳报》在英国大选前支持斯塔默领导的工党\",\n        \"body\": \"鲁珀特·默多克旗下的《太阳报》宣布支持基尔·斯塔默及其反对党工党赢得英国大选，这一戏剧性的举动标志着英国媒体格局的重大转变，也反映了该国政治版图的变化。\",\n        \"url\": \"https:\u002F\u002Fwww.msn.com\u002Fen-us\u002Fmoney\u002Fother\u002Fmurdoch-s-sun-endorses-starmer-s-labour-day-before-uk-vote\u002Far-BB1plQwl\",\n        \"image\": \"https:\u002F\u002Fimg-s-msn-com.akamaized.net\u002Ftenant\u002Famp\u002Fentityid\u002FBB1plZil.img?w=2000&h=1333&m=4&q=79\",\n        \"source\": \"彭博社在MSN.com上\",\n    }, ...\n]\n```\n\n[返回顶部](#TOP)\n\n## 5. books()\n\n```python\ndef books(\n    query: str,\n    max_results: int | None = 10,\n    page: int = 1,\n    backend: str = \"auto\",\n) -> list[dict[str, str]]:\n    \"\"\"DDGS 图书元搜索引擎。\n\n    参数:\n        query: 图书搜索查询。\n        max_results: 最大结果数。默认为10。\n        page: 结果页码。默认为1。\n        backend: 单个或逗号分隔的后端。默认为“auto”。\n\n    返回:\n        包含图书搜索结果的字典列表。\n    \"\"\"\n```\n***示例***\n```python\nresults = DDGS().books(query=\"海狼 杰克·伦敦\", page=1, backend=\"auto\")\nprint(results)\n[\n    {\n        'title': '海狼',\n        'author': '杰克·伦敦',\n        'publisher': 'DigiCat, 2022',\n        'info': '英语 [en], .epub, 🚀\u002Fzlib, 0.5MB, 📗 书籍（未知）',\n        'url': 'https:\u002F\u002Fannas-archive.li\u002Fmd5\u002F574f6556f1df6717de4044e36c7c2782',\n        'thumbnail': 'https:\u002F\u002Fs3proxy.cdn-zlib.sk\u002F\u002Fcovers299\u002Fcollections\u002Fuserbooks\u002Fda4954486be7c2b2b9f70b2aa5bcf01292de3ea510b5656f892821950ded9ada.jpg',\n    }, ...\n]\n```\n\n[返回顶部](#TOP)\n\n## 6. extract()\n\n获取指定 URL 的内容，并以多种格式提取。\n\n```python\ndef extract(\n    url: str,\n    fmt: str = \"text_markdown\",\n) -> dict[str, str | bytes]:\n    \"\"\"获取 URL 并提取其内容。\n\n    参数:\n        url: 要获取并提取内容的 URL。\n        fmt: 输出格式：\n            \"text_markdown\"（HTML→Markdown，保留链接\u002F标题\u002F列表），\n            \"text_plain\"（HTML→纯文本），\n            \"text_rich\"（HTML→带标题\u002F列表的富文本），\n            \"text\"（原始 HTML），\n            \"content\"（原始字节数据）。\n\n    返回:\n        包含 'url' 和 'content' 键的字典。\n    \"\"\"\n```\n***示例***\n```python\n# Markdown（默认） - 保留链接、标题、列表\nresult = DDGS().extract(\"https:\u002F\u002Fexample.com\")\nprint(result)\n{\"url\": \"https:\u002F\u002Fexample.com\", \"content\": \"# 示例域名\\n\\n此域名用于...\"}\n\n# 纯文本\nresult = DDGS().extract(\"https:\u002F\u002Fexample.com\", fmt=\"text_plain\")\n\n# 富文本（标题\u002F列表，无链接 URL）\nresult = DDGS().extract(\"https:\u002F\u002Fexample.com\", fmt=\"text_rich\")\n\n# 原始 HTML\nresult = DDGS().extract(\"https:\u002F\u002Fexample.com\", fmt=\"text\")\n\n# 原始字节数据\nresult = DDGS().extract(\"https:\u002F\u002Fexample.com\", fmt=\"content\")\n```\n\n***命令行界面***\n```bash\nddgs extract -u https:\u002F\u002Fexample.com\nddgs extract -u https:\u002F\u002Fexample.com -f text_plain\nddgs extract -u https:\u002F\u002Fexample.com -f content -o output.json\n```\n\n[返回顶部](#TOP)\n\n## 免责声明\n\n本库仅用于教育目的。","# DDGS 快速上手指南\n\nDDGS (Dux Distributed Global Search) 是一个聚合了多种网络搜索服务结果的元搜索引擎库，支持文本、图片、视频、新闻和书籍搜索，并提供 CLI、API Server 和 MCP Server 等多种使用方式。\n\n## 环境准备\n\n*   **操作系统**：Linux, macOS, Windows\n*   **Python 版本**：>= 3.10\n*   **前置依赖**：无特殊系统级依赖，需确保 `pip` 可用。\n\n> **提示**：国内用户建议使用国内镜像源加速安装（如清华源、阿里源）。\n\n## 安装步骤\n\n### 1. 基础安装（Python 库）\n仅用于在 Python 代码中调用：\n```bash\npip install -U ddgs -i https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple\n```\n\n### 2. 安装 API 服务器\n如需启动 HTTP API 服务（基于 FastAPI）：\n```bash\npip install -U \"ddgs[api]\" -i https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple\n```\n\n### 3. 安装 MCP 服务器\n如需对接 Cursor、Claude Desktop 等 MCP 客户端：\n```bash\npip install -U \"ddgs[mcp]\" -i https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple\n```\n\n## 基本使用\n\n### 方式一：Python 代码调用（推荐）\n\n这是最直接的用法，支持文本、图片、视频等多种搜索类型。\n\n**示例：执行文本搜索**\n\n```python\nfrom ddgs import DDGS\n\n# 初始化并执行搜索\n# query: 搜索关键词\n# max_results: 最大返回结果数\n# backend: 指定后端引擎 (如 'google', 'bing', 'duckduckgo')，默认为 'auto' 自动选择\nresults = DDGS().text(\"python programming\", max_results=5, backend=\"auto\")\n\n# 打印结果\nfor item in results:\n    print(f\"标题：{item['title']}\")\n    print(f\"链接：{item['href']}\")\n    print(f\"摘要：{item['body']}\\n\")\n```\n\n**其他搜索类型示例：**\n\n```python\n# 图片搜索\nimages = DDGS().images(\"butterfly\", max_results=3)\n\n# 新闻搜索\nnews = DDGS().news(\"AI technology\", timelimit=\"w\") # 限制为一周内\n\n# 视频搜索\nvideos = DDGS().videos(\"tutorial\", resolution=\"high\")\n```\n\n### 方式二：命令行工具 (CLI)\n\n安装后可直接在终端使用：\n\n```bash\n# 查看帮助\nddgs --help\n\n# 执行文本搜索 (示例)\nddgs text \"python programming\" --max-results 5\n```\n\n### 方式三：启动 API 服务\n\n安装 `ddgs[api]` 后，可启动本地搜索 API 服务：\n\n```bash\n# 前台启动\nddgs api\n\n# 自定义端口启动\nddgs api --host 127.0.0.1 --port 9000\n\n# 启动后可访问 http:\u002F\u002F127.0.0.1:9000\u002Fdocs 查看 Swagger 文档\n```\n\n### 方式四：配置 MCP 服务\n\n适用于 AI 编辑器（如 Cursor）集成。在客户端配置文件中添加：\n\n```json\n{\n  \"mcpServers\": {\n    \"ddgs\": {\n      \"command\": \"ddgs\",\n      \"args\": [\"mcp\"]\n    }\n  }\n}\n```\n\n配置完成后，AI 助手即可直接调用 `search_text`, `search_images` 等工具进行联网搜索。","某科技公司的数据分析师需要快速收集全球多家媒体关于“生成式 AI 监管政策”的最新报道、相关图片及深度书籍，以撰写一份紧急行业简报。\n\n### 没有 ddgs 时\n- **多源切换繁琐**：分析师需手动在 Google、Bing、DuckDuckGo 等多个搜索引擎间反复切换查询，耗时且容易遗漏关键信息。\n- **数据格式混乱**：从不同网站复制的新闻标题、摘要和图片链接格式不统一，后续清洗和整理数据花费了大量精力。\n- **自动化集成困难**：现有的 Python 爬虫脚本针对单一搜索引擎编写，难以灵活扩展至其他引擎，且极易因反爬机制失效。\n- **内容提取低效**：找到有价值的文章链接后，还需单独编写代码或手动打开页面提取正文内容，工作流被频繁打断。\n\n### 使用 ddgs 后\n- **一键聚合搜索**：通过 `DDGS().text()` 单次调用即可同时获取 Bing、Brave、Google 等九个后端引擎的搜索结果，全面覆盖信息源。\n- **结构化数据输出**：直接返回统一的字典列表格式，包含标题、链接、摘要等字段，无需额外清洗即可导入 Pandas 进行分析。\n- **灵活的多模态支持**：利用 `images()` 和 `books()` 方法快速获取配图素材和权威书籍参考，并通过 `extract()` 直接抓取全文内容。\n- **无缝嵌入工作流**：作为标准 Python 库或通过 MCP 服务器集成到 Cursor 等 AI 编辑器中，让自动化情报收集脚本稳定运行。\n\nddgs 将分散的全球搜索能力整合为统一的编程接口，极大提升了多源信息采集的效率与标准化程度。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fdeedy5_ddgs_6b8199ad.png","deedy5",null,"https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fdeedy5_292bfb45.png","https:\u002F\u002Fgithub.com\u002Fdeedy5",[78,82,86,89],{"name":79,"color":80,"percentage":81},"Python","#3572A5",98,{"name":83,"color":84,"percentage":85},"Makefile","#427819",0.7,{"name":87,"color":88,"percentage":85},"Dockerfile","#384d54",{"name":90,"color":91,"percentage":92},"Shell","#89e051",0.6,2396,235,"2026-04-06T16:22:30","MIT",1,"Linux, macOS, Windows","未说明",{"notes":101,"python":102,"dependencies":103},"该工具是一个聚合多种搜索引擎结果的元搜索库，主要依赖网络连接而非本地计算资源。支持通过 pip 安装基础版、API 服务器版（需 FastAPI）或 MCP 服务器版。支持配置 HTTP\u002FSOCKS5 代理。无需 GPU 加速，对内存无特殊高要求，适合常规开发环境运行。",">=3.10",[104,105],"FastAPI (可选，用于 API 服务器)","MCP (可选，用于 MCP 服务器)",[52,16,14,107],"其他",[109,110,111,64,112,113,114,115],"python","search","metasearch","mcp","mcp-server","websearch","api","2026-03-27T02:49:30.150509","2026-04-07T06:14:03.908235",[119,124,129,134,139,144],{"id":120,"question_zh":121,"answer_zh":122,"source_url":123},20986,"在非主线程中运行 ddgs 时出现 'RuntimeError: There is no current event loop' 错误怎么办？","这是一个已知的异步事件循环问题。维护者建议尝试安装特定版本 v4.4.1 来规避此问题，命令为：pip install duckduckgo_search==4.4.1。此外，该问题可能与 nest_asyncio 和 curl-cffi 的兼容性有关，更新到更新的版本可能也已修复此问题。","https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fissues\u002F185",{"id":125,"question_zh":126,"answer_zh":127,"source_url":128},20987,"使用 ddg_images 搜索图片时遇到 'IndexError: list index out of range' 或返回空列表怎么办？","这通常是因为请求过于频繁导致被 DuckDuckGo 限流或封锁 IP。维护者建议尝试使用代理（proxy）来解决此问题。如果使用的是 Python 3.7，请升级 duckduckgo_search 到 v2.9.5 版本。","https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fissues\u002F57",{"id":130,"question_zh":131,"answer_zh":132,"source_url":133},20988,"在使用 Tor 代理时遇到 'KeyError: b'socks5h'' 错误如何解决？","请尝试将库升级到最新版本（如 v9.9.1 或更高），该版本可能已经修复了与 socks5h 协议相关的解析错误。命令：pip install -U duckduckgo_search。","https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fissues\u002F389",{"id":135,"question_zh":136,"answer_zh":137,"source_url":138},20989,"遇到 '202 Ratelimit' 错误且更新到最新版仍未解决怎么办？","当 max_results 设置过大（例如超过 10）时容易触发此限流错误。请确保升级到至少 v8.0.2 版本，该版本针对限流问题进行了修复。如果问题依旧，请减少单次请求的结果数量或增加请求间隔。","https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fissues\u002F304",{"id":140,"question_zh":141,"answer_zh":142,"source_url":143},20990,"ddg_images() 函数第一次运行正常，第二次运行却返回空数组是什么原因？","这是因为短时间内频繁请求导致 IP 被临时封锁。维护者建议升级到 v3.0.2 或更高版本以改善此情况。命令：pip install -U duckduckgo_search。如果是在 Kaggle 等环境中，等待约 10 分钟让 IP 变更也可能恢复。","https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fissues\u002F60",{"id":145,"question_zh":146,"answer_zh":147,"source_url":133},20991,"如何在代码中正确初始化 DDGS 以避免常见的连接或配置错误？","确保在初始化 DDGS 时正确传递参数。如果遇到后端特定错误（如 duckduckgo 后端报错而 Bing 正常），请检查是否使用了过时的代理配置或旧版本库。通常升级到最新版本并移除不必要的自定义代理设置（除非确实需要 Tor 等）可以解决问题。",[149,154,159,164,169,174,179,184,189,194,199,204,209,214,219,224,229,234,239,244],{"id":150,"version":151,"summary_zh":152,"released_at":153},127014,"v9.13.0","## 变更内容\n* 功能（引擎）：由 @deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F446 中添加了 BingImages 引擎\n* 重构（API）：由 @deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F447 中将 MCP 服务器与 FastAPI 解耦\n\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.12.1...v9.13.0","2026-04-06T15:00:15",{"id":155,"version":156,"summary_zh":157,"released_at":158},127015,"v9.12.1","## 变更内容\n* 重构（ddgs）：由 @deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F445 中将全局的 ThreadPoolExecutor 替换为每个调用独立的有界执行器。\n\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.12.0...v9.12.1","2026-04-03T09:38:27",{"id":160,"version":161,"summary_zh":162,"released_at":163},127016,"v9.12.0","## 变更内容\n* 修复（Google）：使用 Android 用户代理，由 @unixfox 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F438 中实现\n* 新特性：添加通过 extract() 方法提取 URL 内容的功能，由 @deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F440 中实现\n* 新特性：添加 STDIO mcp 端点；新增 AGENTS.md 和 SKILLS.md 文件，由 @deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F441 中实现\n\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.11.4...v9.12.0","2026-03-27T16:15:41",{"id":165,"version":166,"summary_zh":167,"released_at":168},127017,"v9.11.4","## 变更内容\n* 修复（Google）：使用 GSA 用户代理，由 @unixfox 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F428 中完成\n\n## 新贡献者\n* @unixfox 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F428 中完成了首次贡献\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.11.3...v9.11.4","2026-03-14T18:14:58",{"id":170,"version":171,"summary_zh":172,"released_at":173},127018,"v9.11.3","## 变更内容\n\n* [chore(deps): 将 primp 从 1.1.2 升级到 1.1.3](https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcommit\u002F03a18bbb7a0331543ee0447ebf35084ba46e58e0)\n* [refactor(annasarchive): 随机化搜索域名选择](https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcommit\u002Fbf8670d0e9628a09e38317ab71b48167fed3eb5e)\n","2026-03-11T07:11:39",{"id":175,"version":176,"summary_zh":177,"released_at":178},127019,"v9.10.0","## 变更内容\n* 搜索引擎：添加由 @deedy5 提供的 Grokipedia，详见 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F404\n* 搜索引擎：修复并启用 Google 搜索引擎，由 @deedy5 完成，详见 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F405\n\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.9.3...v9.10.0","2025-12-17T23:29:51",{"id":180,"version":181,"summary_zh":182,"released_at":183},127020,"v9.9.3","## 变更内容\n* 暴露强类型化的 DDGS 别名，以修复类型检查器的查找错误，由 @deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F402 中完成\n* CLI：添加 `--no-color` 选项，用于禁用彩色输出，由 @deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F403 中完成\n\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.9.2...v9.9.3","2025-12-05T12:24:50",{"id":185,"version":186,"summary_zh":187,"released_at":188},127021,"v9.9.2","## 变更内容\n* 修复(Brave)：@deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F397 中更新了 XPath 表达式\n* 修复(DuckDuckGo)：@deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F398 中添加了广告过滤功能\n* 修复(Bing)：由于搜索结果不相关，已禁用该引擎，相关更改由 @deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F399 中完成\n* 杂项：@deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F400 中更新了贡献指南\n\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.9.1...v9.9.2","2025-11-29T13:45:12",{"id":190,"version":191,"summary_zh":192,"released_at":193},127022,"v9.9.1","## 变更内容\n* 由 @deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F386 中强制执行 ruff 规则\n* 由 @deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F387 中添加 pre-commit 钩子\n* 修复（DuckDuckGo）：由 @deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F391 中为 HTTP 客户端添加虚假的 User-Agent 和请求头\n\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.9.0...v9.9.1","2025-11-14T16:03:27",{"id":195,"version":196,"summary_zh":197,"released_at":198},127023,"v9.9.0","## 变更内容\n* 移除 orjson，由 @deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F379 中完成\n* 添加 Makefile，由 @deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F380 中完成\n* 为兼容性降级 lxml，由 @deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F381 中完成\n* 引擎：移除 MullvadLetaBrave 和 MullvadLetaGoogle，由 @deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F382 中完成\n* 移除对 Python 3.9 的支持，由 @deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F383 中完成\n* 强制执行更多 ruff 规则：FA、ISC、PIE、PERF、PLC、PLE，由 @deedy5 在 https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F384 中完成\n\n\n**完整变更日志**：https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.8.0...v9.9.0","2025-11-09T15:16:13",{"id":200,"version":201,"summary_zh":202,"released_at":203},127024,"v9.8.0","## What's Changed\r\n* Allow custom CA path via DDGS(verify=) by @deedy5 in https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F378\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.7.1...v9.8.0","2025-11-07T05:02:16",{"id":205,"version":206,"summary_zh":207,"released_at":208},127025,"v9.7.1","## What's Changed\r\n* Fix(typing): replace _DDGSLazyLoader with metaclass-based _ProxyMeta by @deedy5 in https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F377\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.7.0...v9.7.1","2025-11-05T08:15:11",{"id":210,"version":211,"summary_zh":212,"released_at":213},127026,"v9.7.0","## What's Changed\r\n* Make DDGS lazy-loaded by @deedy5 in https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F374\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.6.1...v9.7.0","2025-11-02T15:27:12",{"id":215,"version":216,"summary_zh":217,"released_at":218},127027,"v9.6.1","[fix: disable Google engine](https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcommit\u002F4da7aa84b09ec625575303615c79099a7127232c)\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.6.0...v9.6.1","2025-10-12T18:36:12",{"id":220,"version":221,"summary_zh":222,"released_at":223},127028,"v9.6.0","## What's Changed\r\n* Implement BingNews backend by @deedy5 in https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F372\r\n* Fix(Duckduckgo): add patched http_client2 by @deedy5 in https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F373\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.5.5...v9.6.0","2025-09-17T13:26:47",{"id":225,"version":226,"summary_zh":227,"released_at":228},127029,"v9.5.5","## What's Changed\r\n* CLI: enable warning-level logging by @deedy5 in https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F369\r\n* CLI: enable warnings logging in safe_entry_point by @deedy5 in https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F370\r\n* Fix AnnasArchive backend: update xpath's by @deedy5 in https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F371\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.5.4...v9.5.5","2025-09-01T13:24:47",{"id":230,"version":231,"summary_zh":232,"released_at":233},127030,"v9.5.4","## What's Changed\r\n* Bugfix: parsing from bytes causes encoding issues (revert \"perf: zero-copy parsing tree from html memoryview\" ) by @deedy5 in https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F367\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.5.3...v9.5.4","2025-08-17T09:57:37",{"id":235,"version":236,"summary_zh":237,"released_at":238},127031,"v9.5.3","## What's Changed\r\n* Improve code quality by enabling Ruff \"D\" (pydocstyle) rules and updating docstrings by @deedy5 in https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F363\r\n* Improve code quality by enabling Ruff \"S\" (flake8-bandit) rules by @deedy5 in https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F364\r\n* Improve code quality by enabling Ruff \"C4\", \"RUF\", \"EM\", \"RET\", \"LOG\", \"G\", \"DTZ\" rules by @deedy5 in https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F365\r\n* Zero-copy parsing tree from html memoryview  by @deedy5 in https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F366\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.5.2...v9.5.3","2025-08-17T01:43:48",{"id":240,"version":241,"summary_zh":242,"released_at":243},127032,"v9.5.2","## What's Changed\r\n* backend: don't include wikipedia if not `auto` by @deedy5 in https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F359\r\n* Fix Mojeek backend: use cookies to set country ang lang params by @deedy5 in https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F361\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.5.1...v9.5.2","2025-08-05T21:11:28",{"id":245,"version":246,"summary_zh":247,"released_at":248},127033,"v9.5.1","## What's Changed\r\n* Fix Mojeek engine, enable Bing engine by @deedy5 in https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fpull\u002F355\r\n\r\n\r\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fdeedy5\u002Fddgs\u002Fcompare\u002Fv9.5.0...v9.5.1","2025-08-01T10:09:42"]