How openzim/kolibri works
Main Components
a. Input Layer:
• Kolibri Channel Interface (handles channel ID and node IDs)
• Configuration Handler (processes command-line arguments)
b. Core Processing Components:
• Scraper Module (downloads content from Kolibri)
• Content Processors:
• Video Processor (handles webm/mp4)
• Document Processor (handles PDFs/ePubs)
• Audio Processor
• Subtitle Handler
• Thumbnail Processor
c. Frontend UI (Vue.js based):
• Topic Navigation System
• Content Viewer
• Search Interface
d. ZIM Creation Layer:
• Content Organizer
• Static HTML Generator
• ZIM File Builder
[Input Layer]
• ↓
[Processing Layer]
• ↓
[Frontend Layer]
• ↓
[Output Layer]
Key Relationships to Highlight
• Data flow from Kolibri Channel input through the processing pipeline
• How the Vue.js frontend (zimui) integrates with the processed content
• Demonstrate the final packaging into ZIM format
Technical Details
a. Technologies:
• Python 3.11 (main scraper)
• Node.js 20.x (frontend)
• FFmpeg (media processing)
• Vue.js (UI framework)
• TypeScript (frontend type system)
b. Important Files/Components:
• scraper/src/kolibri2zim/scraper.py (main scraping logic)
• scraper/src/kolibri2zim/processing.py (content processing)
• zimui/src/components/* (UI components)
Specific Implementation Notes
• The separation between the Python scraper and Vue.js frontend
• The content processing pipeline's modularity
• How different content types (video, audio, documents) are handled
• The final ZIM packaging process