@sap-ux/fiori-docs-embeddings

A Node.js package that indexes SAP Fiori related documentation from multiple sources and generates vector embeddings for semantic search capabilities.

What it does

This module- Crawls documentation from GitHub repositories (see Configuration for a list of default sources) Example sources:

https://github.com/SAP-docs/btp-fiori-tools (Fiori Tools documentation)

https://github.com/SAP-docs/sapui5/tree/main/docs/06_SAP_Fiori_Elements (UI5 Fiori Elements documentation)

https://github.com/SAP-samples/fiori-tools-samples (Sample applications)

https://github.com/SAP-samples/fiori-elements-feature-showcase (Feature examples)

https://ui5.sap.com/test-resources/sap/fe/macros/designtime/apiref/api.json (Design Time API Reference)

https://ui5.sap.com/test-resources/sap/fe/core/fpmExplorer/index.html#/overview/introduction SAP Fiori Development Portal

• Parses markdown, JSON, TypeScript, and other file types
• Generates AI-powered vector embeddings using transformers
• Stores embeddings in a local LanceDB vector database
• Provides tools for semantic and keyword search across documentation

Installation

npm install @sap-ux/fiori-docs-embeddings

Usage

Basic Usage

import { getDataPath, getEmbeddingsPath } from '@sap-ux/fiori-docs-embeddings';

// Get paths to data directories
const docsPath = getDataPath();
const embeddingsPath = getEmbeddingsPath();

Building Documentation Index

# Set GitHub token to avoid rate limits
export GITHUB_TOKEN=your_github_token

# Build documentation index
npm run update-docs

# Generate embeddings
npm run update-embeddings

# Or do both
npm run update-all

Available Scripts

• update-docs - Crawl and index documentation from configured sources
• update-embeddings - Generate vector embeddings from indexed documents
• update-all - Run both documentation indexing and embedding generation

Configuration

The module indexes documentation from these sources by default:

• SAP-docs/btp-fiori-tools (Fiori Tools documentation)
• SAP-docs/sapui5 (UI5 framework documentation)
• SAP-samples/fiori-tools-samples (Sample applications)
• SAP-samples/fiori-elements-feature-showcase (Feature examples)

Environment Variables

• GITHUB_TOKEN - GitHub personal access token (recommended to avoid rate limits)

Data Structure

Generated data is organized as:

data/
├── docs/           # Parsed documentation files
├── embeddings/     # Vector database (LanceDB)
└── search/         # Search indexes

Features

• Multi-source indexing - Supports GitHub repositories and JSON APIs
• File type support - Markdown, JSON, TypeScript, JavaScript, XML, YAML, and more
• Vector embeddings - Uses sentence-transformers/all-MiniLM-L6-v2 model
• Local storage - All data stored locally with LanceDB
• Caching - Intelligent caching to avoid unnecessary API calls
• Chunking - Smart document chunking for optimal embedding generation

Development

# Install dependencies
npm install

# Run tests
npm test

# Check types
npm run build-compiler

# Clean generated files
npm run clean

License

Apache-2.0