English | 中文

@1-/md : Extract code block positions and languages from Markdown

• @1-/md : Extract code block positions and languages from Markdown
  • 1. Features
  • 2. Usage
  • 3. Design
  • 4. Tech Stack
  • 5. Code Structure
  • 6. History
  • About

1. Features

Extract code block locations and language identifiers from Markdown text.

• Normalize cross-platform line endings (\r\n, \r) to standard \n
• Support fenced code blocks with any number of consecutive backticks (≥3)
• Accurately extract language identifiers, including empty strings for language-agnostic blocks
• Record start and end line numbers (1-based indexing) for each code block
• Automatically close unclosed code blocks at document end

2. Usage

import li from "@1-/md/li.js";
import code from "@1-/md/code.js";

const markdownContent = `# Title

\`\`\`javascript
const val = 1;
\`\`\``;

// Split Markdown text into normalized line array
const lines = li(markdownContent);

// Extract all code block information
const blocks = code(lines);

console.log(blocks);
// Output format: [ [ Language, Start Line, End Line ] ]
// Example output: [ [ 'javascript', 3, 5 ] ]

3. Design

The system consists of two independent modules: line normalization (li.js) and code block extraction (code.js).

li.js normalizes cross-platform line endings and splits text into line arrays while trimming trailing whitespace from each line.

code.js uses a state machine to iterate through the line array:

• Initial state (outside code block): match lines starting with ≥3 backticks, record backtick count, language, and start line number, transition to inside state
• Inside code block state: match ending lines with identical backtick count, record end line number and save block information, transition back to outside state
• Document end: if still inside a code block, use total line count as end line number and save the unclosed block

4. Tech Stack

• Runtime: Bun / Node.js
• Language: JavaScript (ES Modules)
• Linter: Oxlint
• Formatter: Oxfmt

5. Code Structure

.
├── src/
│   ├── code.js          # Code block extraction logic
│   └── li.js            # Line normalization processing
└── test/
    ├── _.test.js        # Unit test
    └── test.md          # Markdown test cases

6. History

In 2004, John Gruber and Aaron Swartz co-created the Markdown markup language, with early specifications supporting only indented code blocks.

In 2012, GitHub introduced fenced code blocks in GitHub Flavored Markdown (GFM), using backticks (`) to wrap code with optional language identifiers, significantly improving readability and practicality of technical documentation.

In 2014, fenced code blocks were standardized in the CommonMark specification, becoming a core feature of modern Markdown implementations across documentation generators, static site builders, and development tools.

About

This library is developed by WebC.site.

WebC.site: A new paradigm of web development for AI

@1-/md : 解析 Markdown 提取代码块位置与语言类型

• @1-/md : 解析 Markdown 提取代码块位置与语言类型
  • 1. 功能介绍
  • 2. 使用演示
  • 3. 设计思路
  • 4. 技术栈
  • 5. 代码结构
  • 6. 历史故事
  • 关于

1. 功能介绍

提取 Markdown 文本中围栏代码块的位置信息与语言标识。

• 统一处理跨平台换行符（\r\n、\r）为标准 \n
• 支持任意数量连续反引号（≥3）定义的围栏代码块
• 准确提取代码块语言标识，包括无语言标识的空字符串情况
• 记录每个代码块的起始行号与结束行号（1-based indexing）
• 自动闭合文档末尾未配对的代码块

2. 使用演示

import li from "@1-/md/li.js";
import code from "@1-/md/code.js";

const markdownContent = `# Title

\`\`\`javascript
const val = 1;
\`\`\``;

// 将 Markdown 文本拆分为标准化的行数组
const lines = li(markdownContent);

// 提取所有代码块信息
const blocks = code(lines);

console.log(blocks);
// 输出格式: [ [ 语言类型, 起始行号, 结束行号 ] ]
// 示例输出: [ [ 'javascript', 3, 5 ] ]

3. 设计思路

系统由两个独立模块组成：行标准化模块（li.js）和代码块提取模块（code.js）。

li.js 模块统一处理不同平台的换行符，将文本分割为行数组，并清除每行末尾空白字符。

code.js 模块使用状态机遍历行数组：

• 初始状态（代码块外）：匹配以 ≥3 个反引号开头的行，记录反引号数量、语言类型及起始行号，切换至代码块内状态
• 代码块内状态：匹配相同数量反引号的结束行，记录结束行号并保存代码块信息，切换回代码块外状态
• 文档末尾：若仍处于代码块内状态，则以文档总行数作为结束行号，保存未闭合代码块

4. 技术栈

• 运行时: Bun / Node.js
• 语言: JavaScript (ES Modules)
• 代码检查: Oxlint
• 代码格式化: Oxfmt

5. 代码结构

.
├── src/
│   ├── code.js          # 代码块提取逻辑
│   └── li.js            # 行标准化处理
└── test/
    ├── _.test.js        # 单元测试
    └── test.md          # 测试用 Markdown 样例

6. 历史故事

2004 年，John Gruber 与 Aaron Swartz 共同创建 Markdown 标记语言，早期规范仅支持缩进方式定义代码块。

2012 年，GitHub 在 GitHub Flavored Markdown (GFM) 中引入围栏代码块语法，使用反引号（`）包裹代码并支持语言标识，大幅提升技术文档可读性与实用性。

2014 年，围栏代码块被纳入 CommonMark 规范，成为现代 Markdown 的标准特性，广泛应用于文档生成、静态网站构建和开发工具中。

关于

本库由 WebC.site 开发。

WebC.site : 面向人工智能的网站开发新范式