slipbot-logseq-importer（slipbot logeq 导入器）

Logseq Import

Parse a Logseq page and create individual slipbox entries for each bullet point.

Critical Rule: Ignore All Tags

Do not import any tags from Logseq. This includes:

• Page-level tags:: property
• Inline #tags in bullet content
• block-tags:: metadata

Slipbot generates its own tags based on content. Logseq tags would conflict with this.

Input Format

Logseq pages have two parts:

1. Page Properties (top of page, key:: value format):

type:: #literature
source:: Book
author:: David Kadavy
title:: Digital Zettelkasten
alias:: zettelkasten-book
status::
tags::

2. Bulleted Notes (markdown list):

- First note content here
- Second note with [[page ref]] link
  - Nested bullet under second
- Third note id:: abc123-uuid

Property Mapping

Empty properties (e.g., author:: with no value) → null

Parsing Rules

Properties

1. Extract all key:: value lines at the top
2. Stop when hitting first bullet (- )
3. Strip # from values like #literature
4. Parse markdown links: [text](url) → extract both parts

Bullets

1. Each top-level bullet (- ) becomes its own slipbox note
2. Nested bullets: Add parent context to make them standalone

- Example parent: - [[Fleeting Notes]]: quick notes written anywhere - Example child: - Can be on paper or digital - Result: "Fleeting Notes (quick notes written anywhere) can be on paper or digital"

1. Strip Logseq metadata from bullets:

- id:: uuid → remove - block-tags:: #xxx → remove entirely - #tag inline tags → remove entirely (slipbot generates its own) - #{"{" or malformed tags → remove

1. Convert [[page refs]] → plain text (potential link targets)

Content Cleanup

• Remove trailing id:: xxx from bullets
• Remove block-tags:: xxx entirely
• Remove all #tag inline tags (slipbot generates its own tags)
• Preserve markdown formatting (bold, italic, code)

Workflow

1. Precheck (before import)

- Parse the page properties and bullets (don't create notes yet) - Generate a brief summary of what the page is about (1-2 sentences based on title, author, and content themes) - Count total notes that will be created (including nested bullets that become standalone) - Present to user: summary, note count, source info - Ask for confirmation before proceeding with import - If user declines, stop and don't create any notes

1. Parse the page (after confirmation)

- Extract page properties → source metadata - Extract all bullets → note list - Handle nesting by enriching child bullets with parent context

1. For each bullet, invoke slipbot workflow:

- Use - {content} prefix (note type) - Include source: ~ {source.type}, {source.title} by {source.author} - Let slipbot handle: filename, tags, links, graph update

1. Report results

- Count of notes created - Any issues encountered

Example

Input:

type:: #literature
source:: Book
author:: David Kadavy
title:: Digital Zettelkasten

- Rewriting ideas helps decide their importance
- [[Fleeting Notes]]: quick notes written anywhere
  - Can be captured on paper or digitally
- Keywords should be specific to the idea id:: abc123

Processing:

1. Source: Book, "Digital Zettelkasten" by David Kadavy
2. Bullets extracted:

- "Rewriting ideas helps decide their importance" - "Fleeting Notes: quick notes written anywhere" - "Fleeting Notes can be captured on paper or digitally" (nested, parent context added) - "Keywords should be specific to the idea" (id stripped)

1. Each sent to slipbot as:

- - Rewriting ideas helps decide their importance ~ Book, Digital Zettelkasten by David Kadavy - etc.

Edge Cases

• No properties: Treat as standalone notes without source
• Deeply nested bullets (3+ levels): Flatten, accumulating context from all ancestors
• Citation sections (## Citation:): Ignore (redundant with properties)
• Non-bullet content: Ignore headers, paragraphs outside bullets
• Multiple [[refs]] in one bullet: Keep all, convert to plain text