* Update chunking-and-cleaning-text.mdx 原文中 '通用模式' 小节中,关于 '分段标识符' 的解释中描述如下: `分段标识符,默认值为 \n,即按照文章段落进行分块。你可以遵循正则表达式语法自定义分块规则,系统将在文本出现分段标识符时自动执行分段。例如 的含义是按照句子进行分段。下图是不同语法的文本分段效果:` `例如` 后好像缺少了转义字符,这里补充正则表达式 `(?<=[.!?])\s+` 以查找句子结束标点(., !, ?)后的空白字符,从而按照句子分段。 此外,在 "父子模式" 小节中,以下内容好像也遗漏了转义字符: `在子分段内填写以下分段设置: 分段标识符,默认值为 ,即按照句子进行分段。你可以遵循正则表达式语法自定义分块规则,系统将在文本出现分段标识符时自动执行分段。` 这里补充 `\.|\!|\?` 作为按句子分段的标识符,供作者审核。 * correct, update, and remove expired content --------- Co-authored-by: Riskey <riskey47@dify.ai>
📘 Dify Documentation
Welcome to the documentation repository for Dify.
We warmly welcome your contributions — whether it’s proofreading, fixing typos, or submitting new content. Please feel free to open issues or PRs if you find anything that could be improved!
🚀 Project Overview
This project uses the Mintlify Kit to build and serve modern, developer-friendly documentation.
🛠️ Local Development
To preview and develop documentation locally:
-
Install the Mintlify CLI.
npm i -g mintlify -
Start local development.
Run the following command at the root of your project (where docs.json is located):
mintlify dev
🙌 Contributing
Your help in reviewing, editing, and expanding the documentation is truly appreciated.
📝 Contribution Workflow
-
Fork this repository to your own GitHub account.
-
Create a new branch based on the
mainbranch. -
Start the local development server following the steps above to preview your changes live.
-
Make your edits or write new content in the appropriate file under the content/ directory.
-
Submit a Pull Request (PR) after verifying your changes:
- If you’ve added new pages or sections, don’t forget to update docs.json to include them in the sidebar navigation.
- We welcome tri-lingual contributions (English, Simplified Chinese, Japanese) — contribute in one or more languages if possible.
Please submit PRs to this repository instead of the legacy one: 📘 https://github.com/langgenius/dify-docs-archived
Thanks again for being part of Dify’s documentation journey!
📄 License
The Dify product documentation in the assets, content, and data folders are licensed under a CC-BY license.