Mind Elixir 纯文本格式详解：思维导图数据转换与AI编辑规范

Mind Elixir Plaintext Format by ssshooter/mind-elixir-core

3,000 GitHub Stars

GitHub

安装命令

npx skills add https://github.com/ssshooter/mind-elixir-core --skill 'Mind Elixir Plaintext Format'

数据可视化开发知识管理

🇨🇳中文介绍

Mind Elixir 纯文本格式

此技能解释了 Mind Elixir 纯文本格式规范，以及如何在它与 Mind Elixir JSON 数据结构之间进行转换。纯文本格式非常适合人工编辑、AI 生成和流式传输场景。

1. 格式概述

纯文本格式使用基于缩进的语法，类似于 Markdown 列表。每一行代表思维导图中的一个节点，缩进（每级 2 个空格）定义了层级关系。

基本结构

- Root Node
  - Child Node 1
    - Child Node 1-1
    - Child Node 1-2
  - Child Node 2
    - Child Node 2-1

关键规则：

根节点以 - 开头（短横线后跟一个空格）
每个子节点相对于其父节点缩进 2 个空格
每个节点行以 - 开头，后跟一个空格和主题文本

2. 高级功能

2.1 节点引用 (ID)

使用 [^id] 语法为节点分配自定义 ID，以便进行交叉引用：

- Root
  - Node A [^id1]
  - Node B [^id2]

格式： [^customId] 放在主题文本的末尾。

广告位招租

在这里展示您的产品或服务

触达数万 AI 开发者，精准高效

联系我们

3.1 纯文本转 Mind Elixir 数据

Mind Elixir 提供了一个内置的转换器用于解析纯文本：

import { plaintextToMindElixir } from 'mind-elixir/plaintextConverter'

const plaintext = `
- Root Node
  - Child 1
  - Child 2
`

const mindElixirData = plaintextToMindElixir(plaintext)
// 返回 MindElixirData 对象，可用于 mind.init() 或 mind.refresh()

3.2 Mind Elixir 数据转纯文本

您也可以将现有的思维导图转换回纯文本格式，这对于导出和保存数据非常有用：

import { mindElixirToPlaintext } from 'mind-elixir/plaintextConverter'

const mindElixirData = mind.getAllData()
const plaintext = mindElixirToPlaintext(mindElixirData)
// 返回包含纯文本格式思维导图的字符串

import MindElixir from 'mind-elixir'
import { plaintextToMindElixir } from 'mind-elixir/plaintextConverter'

// 1. 解析纯文本
const plaintext = `
- My Mind Map
  - Topic 1
    - Subtopic 1.1
    - Subtopic 1.2
  - Topic 2
`

const data = plaintextToMindElixir(plaintext)

// 2. 初始化 Mind Elixir
const mind = new MindElixir({
  el: '#map',
  direction: MindElixir.RIGHT,
})

mind.init(data)

4.1 AI 生成的思维导图

纯文本格式非常适合大语言模型生成：

// 给 AI 的提示示例
const prompt = `
Generate a mind map in plaintext format about "Web Development".
Use this format:
- Root Topic
  - Subtopic 1
    - Detail 1.1
  - Subtopic 2
`

// 然后解析 AI 的响应
const aiResponse = await callAI(prompt)
const data = plaintextToMindElixir(aiResponse)
mind.refresh(data)

4.2 基于文件的思维导图

将思维导图存储为 .txt 或 .md 文件：

// 从文件加载
const response = await fetch('/mindmaps/project.txt')
const plaintext = await response.text()
const data = plaintextToMindElixir(plaintext)
mind.init(data)

纯文本格式对于版本控制来说易于合并：

# 在 Git 中易于 diff 和 merge
git diff mindmap.txt

始终将解析包装在 try-catch 块中：

function safeParse(plaintext: string): MindElixirData | null {
  try {
    return plaintextToMindElixir(plaintext)
  } catch (error) {
    console.error('Parse error:', error)
    // 返回备用数据
    return {
      nodeData: {
        id: 'root',
        topic: 'Parse Error',
        children: []
      }
    }
  }
}

一致的缩进：始终每级使用 2 个空格
唯一 ID：使用引用时，确保 ID 在文档内是唯一的
链接放置：链接可以放在任何地方，但为了可读性，请将其放在相关节点附近
验证：在解析前验证纯文本，特别是对于用户输入
流式传输：对于流式传输场景，增量解析并优雅地处理部分数据

7. 格式规范摘要

功能	语法	示例
节点	`- 主题`	`- My Node`
带 ID 的节点	`- 主题 [^id]`	`- Node A [^id1]`
带样式的节点	`- 主题 {"prop": "value"}`	`- Node {"color": "#ff0000"}`
双向链接	`> [^id1] (x,y) <-标签-> (x,y) [^id2]`	`> [^a] (10,20) <-connects-> (-10,-20) [^b]`
正向链接	`> [^id1] (x,y) >-标签-> (x,y) [^id2]`	`> [^a] (10,20) >-leads to-> (-10,-20) [^b]`
摘要（全部）	`} 摘要文本`	`} Overview`
摘要（N 个节点）	`}:N 摘要文本`	`}:3 Last three items`

8. TypeScript 类型

import type { MindElixirData, NodeObj } from 'mind-elixir'

// 转换器函数
function plaintextToMindElixir(plaintext: string): MindElixirData
function mindElixirToPlaintext(data: MindElixirData): string

🇺🇸English

Mind Elixir Plaintext Format

This skill explains the Mind Elixir plaintext format specification and how to convert between it and the Mind Elixir JSON data structure. The plaintext format is ideal for human editing, AI generation, and streaming scenarios.

1. Format Overview

The plaintext format uses indentation-based syntax similar to Markdown lists. Each line represents a node in the mind map, and indentation (2 spaces per level) defines the hierarchy.

Basic Structure

- Root Node
  - Child Node 1
    - Child Node 1-1
    - Child Node 1-2
  - Child Node 2
    - Child Node 2-1

Key Rules:

Root node starts with - (dash followed by space)
Each child is indented by 2 spaces relative to its parent
Each node line starts with - followed by a space and the topic text

2. Advanced Features

2.1 Node References (IDs)

Assign custom IDs to nodes for cross-referencing using [^id] syntax:

- Root
  - Node A [^id1]
  - Node B [^id2]

Format: [^customId] at the end of the topic text.

2.2 Node Links

Create connections between nodes using the > prefix and arrow syntax:

Bidirectional Links

- Root
  - Node A [^id1]
  - Node B [^id2]
  - > [^id1] (10,20) <-Link Label-> (-10,-20) [^id2]

Format: > [^sourceId] (delta1X,delta1Y) <-Label-> (delta2X,delta2Y) [^targetId]

(delta1X,delta1Y): Offset of the control point from the start node.
(delta2X,delta2Y): Offset of the control point from the end node.

[!TIP] Optional Coordinates : When manually writing or generating plaintext (e.g., via AI), you can omit the (x,y) coordinates. Mind Elixir will automatically calculate default balanced offsets when rendering. However, when you export data back to plaintext, these coordinates will be included to preserve any manual adjustments made in the UI.

Unidirectional Links

- Root
  - Node A [^id1]
  - Node B [^id2]
  - > [^id1] >-Forward Link-> [^id2]

Formats:

Forward: > [^sourceId] >-Label-> [^targetId]

2.3 Node Styling

Apply inline styles using JSON-like syntax:

- Root
  - Styled Node {"color": "#e87a90"}
  - Another Node {"color": "#3298db", "background": "#ecf0f1"}

Format: {"property": "value", "property2": "value2"} at the end of the topic text.

Common Properties:

color: Text color (hex code)
background: Background color (hex code)
fontSize: Font size (number as string, e.g., "16")

2.4 Summary Nodes

Create summary nodes that visually group previous siblings:

- Root
  - Node 1
  - Node 2
  - Node 3
  - } Summary of all above nodes

Formats:

} Summary text - Summarizes all previous siblings
}:N Summary text - Summarizes the previous N siblings (e.g., }:2 for last 2 nodes)
- Root
  - Node 1
  - Node 2
  - Node 3
  - }:2 Summary of Node 2 and Node 3

2.5 Complete Example

- Project Planning
  - Phase 1: Research [^phase1]
    - Market Analysis {"color": "#3298db"}
    - Competitor Study {"color": "#3298db"}
    - }:2 Research Summary
  - Phase 2: Development [^phase2]
    - Frontend {"color": "#2ecc71"}
    - Backend {"color": "#2ecc71"}
    - Testing {"color": "#f39c12"}
    - } Development Summary
  - Phase 3: Launch [^phase3]
    - Marketing
    - Deployment
  - > [^phase1] (50,0) >-Leads to-> (-50,0) [^phase2]
  - > [^phase2] >-Leads to-> [^phase3]

3. Conversion API

3.1 Plaintext to Mind Elixir Data

Mind Elixir provides a built-in converter for plaintext parsing:

import { plaintextToMindElixir } from 'mind-elixir/plaintextConverter'

const plaintext = `
- Root Node
  - Child 1
  - Child 2
`

const mindElixirData = plaintextToMindElixir(plaintext)
// Returns MindElixirData object ready for mind.init() or mind.refresh()

3.2 Mind Elixir Data to Plaintext

You can also convert an existing mind map back to the plaintext format, which is useful for exporting and saving data:

import { mindElixirToPlaintext } from 'mind-elixir/plaintextConverter'

const mindElixirData = mind.getAllData()
const plaintext = mindElixirToPlaintext(mindElixirData)
// Returns a string containing the mind map in plaintext format

3.3 Integration Example

import MindElixir from 'mind-elixir'
import { plaintextToMindElixir } from 'mind-elixir/plaintextConverter'

// 1. Parse plaintext
const plaintext = `
- My Mind Map
  - Topic 1
    - Subtopic 1.1
    - Subtopic 1.2
  - Topic 2
`

const data = plaintextToMindElixir(plaintext)

// 2. Initialize Mind Elixir
const mind = new MindElixir({
  el: '#map',
  direction: MindElixir.RIGHT,
})

mind.init(data)

4. Use Cases

4.1 AI-Generated Mind Maps

The plaintext format is ideal for LLM generation:

// Prompt example for AI
const prompt = `
Generate a mind map in plaintext format about "Web Development".
Use this format:
- Root Topic
  - Subtopic 1
    - Detail 1.1
  - Subtopic 2
`

// Then parse the AI response
const aiResponse = await callAI(prompt)
const data = plaintextToMindElixir(aiResponse)
mind.refresh(data)

4.2 File-Based Mind Maps

Store mind maps as .txt or .md files:

// Load from file
const response = await fetch('/mindmaps/project.txt')
const plaintext = await response.text()
const data = plaintextToMindElixir(plaintext)
mind.init(data)

4.3 Collaborative Editing

The plaintext format is merge-friendly for version control:

# Easy to diff and merge in Git
git diff mindmap.txt

5. Error Handling

Always wrap parsing in try-catch blocks:

function safeParse(plaintext: string): MindElixirData | null {
  try {
    return plaintextToMindElixir(plaintext)
  } catch (error) {
    console.error('Parse error:', error)
    // Return fallback data
    return {
      nodeData: {
        id: 'root',
        topic: 'Parse Error',
        children: []
      }
    }
  }
}

6. Best Practices

Consistent Indentation : Always use 2 spaces per level
Unique IDs : When using references, ensure IDs are unique within the document
Link Placement : Links can be placed anywhere, but keep them near related nodes for readability
Validation : Validate plaintext before parsing, especially for user input
Streaming : For streaming scenarios, parse incrementally and handle partial data gracefully

7. Format Specification Summary

Feature	Syntax	Example
Node	`- Topic`	`- My Node`
Node with ID	`- Topic [^id]`	`- Node A [^id1]`
Node with Style	`- Topic {"prop": "value"}`	`- Node {"color": "#ff0000"}`
Bidirectional Link	`> [^id1] (x,y) <-Label-> (x,y) [^id2]`

8. TypeScript Types

import type { MindElixirData, NodeObj } from 'mind-elixir'

// Converter functions
function plaintextToMindElixir(plaintext: string): MindElixirData
function mindElixirToPlaintext(data: MindElixirData): string

Weekly Installs

–

Repository

ssshooter/mind-…xir-core

GitHub Stars

3.0K

First Seen

–

Security Audits

Gen Agent Trust HubPass SocketPass SnykPass

React 组合模式指南：Vercel 组件架构最佳实践，提升代码可维护性

109,600 周安装